Authorea

Marvin Ward Jr added file TEL_converge_recession.ipynb about 10 years ago

Commit id: 181fbc0fae806f0939eb2cfef9a91b48553e84a8

deletions | additions

{ "metadata": { "name": "" }, "nbformat": 3, "nbformat_minor": 0, "worksheets": [ { "cells": [ { "cell_type": "heading", "level": 1, "metadata": {}, "source": [ "Tax and Expenditure Limitations in Recessions" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We have already looked at fiscal and economic capacity county-wise convergence over time in *CO_TEL_convergence*. For the APPAM Conference, this message must take on a decidedly business cycle feel. The purpose of this script is to explore whether or not the convergence we have already observed behaves differently at different times. This script will also provide a more detailed comparison of revenue and economic convergence, as well as feature a new method of measuring TEL impact.\n", "\n", "To fulfill these tasks, we need to integrate demographic data with dataset produced in *CO_TEL_convergence* upfront. Once this is done, we will also pull in the assessment base information." ] }, { "cell_type": "code", "collapsed": false, "input": [ "import numpy as np\n", "import pandas as pd\n", "from pandas import Series, DataFrame\n", "from IPython.display import HTML\n", "import pandas.io.data as web\n", "import pysal\n", "\n", "#Establish working directory\n", "workdir='/home/choct155/Google Drive/Dissertation/Data/'\n", "\n", "#Set print width\n", "pd.set_option('line_width', 140)" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 20 }, { "cell_type": "markdown", "metadata": {}, "source": [ "The dataset from *CO_TEL_convergence* contains both fiscal and spatial attributes." ] }, { "cell_type": "code", "collapsed": false, "input": [ "co_sf=pd.read_csv(workdir+'CO_sf.csv')\n", "\n", "# for i in range(21):\n", "# print co_sf[co_sf.columns[i*6:(i+1)*6]].head(),'\\n'" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 21 }, { "cell_type": "markdown", "metadata": {}, "source": [ "The demographic data are taken from the [Colorado County Profile query tool](https://dola.colorado.gov/demog_webapps/psc_parameters.jsf) maintained by the Department of Local Affairs. Note that the data are arranged in long format, with variable and value columns. " ] }, { "cell_type": "code", "collapsed": false, "input": [ "#Read in data\n", "co_dem=pd.read_csv(workdir+'CO_cty_profile.csv',delimiter='\\t')\n", "\n", "#Rename columns\n", "co_dem.columns=['AUDIT_YEAR','FNAME','var','value']\n", "co_dem" ], "language": "python", "metadata": {}, "outputs": [ { "html": [ "

	AUDIT_YEAR	FNAME	var	value
0	2011	Adams	Average Household Size	2.85
1	2011	Adams	Births	7244.00
2	2011	Adams	Deaths	2462.00
3	2011	Adams	Group Quarters Population (July)	4024.00
4	2011	Adams	Household Population (July)	447552.00
5	2011	Adams	Housing Vacancy Rate	3.84
6	2011	Adams	July Population	451576.00
7	2011	Adams	Natural Population Increase	4782.00
8	2011	Adams	Net Building Permits	226.00
9	2011	Adams	Net Migration	3083.00
10	2011	Adams	Population (U.S. Census Bureau)	451443.00
11	2011	Adams	Total Households	157264.00
12	2011	Adams	Total Housing Units	163550.00
13	2011	Adams	Vacant Housing Units	6286.00
14	2011	Alamosa	Average Household Size	2.45
15	2011	Alamosa	Births	234.00
16	2011	Alamosa	Deaths	131.00
17	2011	Alamosa	Group Quarters Population (July)	740.00
18	2011	Alamosa	Household Population (July)	14902.00
19	2011	Alamosa	Housing Vacancy Rate	7.95
20	2011	Alamosa	July Population	15642.00
21	2011	Alamosa	Natural Population Increase	103.00
22	2011	Alamosa	Net Building Permits	8.00
23	2011	Alamosa	Net Migration	65.00
24	2011	Alamosa	Population (U.S. Census Bureau)	15710.00
25	2011	Alamosa	Total Households	6075.00
26	2011	Alamosa	Total Housing Units	6600.00
27	2011	Alamosa	Vacant Housing Units	525.00
28	2011	Arapahoe	Average Household Size	2.53
29	2011	Arapahoe	Births	7721.00
30	2011	Arapahoe	Deaths	3426.00
31	2011	Arapahoe	Group Quarters Population (July)	4894.00
32	2011	Arapahoe	Household Population (July)	579809.00
33	2011	Arapahoe	Housing Vacancy Rate	4.53
34	2011	Arapahoe	July Population	584703.00
35	2011	Arapahoe	Natural Population Increase	4295.00
36	2011	Arapahoe	Net Building Permits	1102.00
37	2011	Arapahoe	Net Migration	5589.00
38	2011	Arapahoe	Population (U.S. Census Bureau)	584948.00
39	2011	Arapahoe	Total Households	229014.00
40	2011	Arapahoe	Total Housing Units	239870.00
41	2011	Arapahoe	Vacant Housing Units	10856.00
42	2011	Archuleta	Average Household Size	2.27
43	2011	Archuleta	Births	137.00
44	2011	Archuleta	Deaths	82.00
45	2011	Archuleta	Group Quarters Population (July)	129.00
46	2011	Archuleta	Household Population (July)	11909.00
47	2011	Archuleta	Housing Vacancy Rate	40.44
48	2011	Archuleta	July Population	12038.00
49	2011	Archuleta	Natural Population Increase	55.00
50	2011	Archuleta	Net Building Permits	36.00
51	2011	Archuleta	Net Migration	-77.00
52	2011	Archuleta	Population (U.S. Census Bureau)	12013.00
53	2011	Archuleta	Total Households	5247.00
54	2011	Archuleta	Total Housing Units	8810.00
55	2011	Archuleta	Vacant Housing Units	3563.00
56	2011	Baca	Average Household Size	2.20
57	2011	Baca	Births	41.00
58	2011	Baca	Deaths	73.00
59	2011	Baca	Group Quarters Population (July)	82.00
	...	...	...	...

\n", "

24192 rows \u00d7 4 columns

\n", "

" ], "metadata": {}, "output_type": "pyout", "prompt_number": 22, "text": [ " AUDIT_YEAR FNAME var value\n", "0 2011 Adams Average Household Size 2.85\n", "1 2011 Adams Births 7244.00\n", "2 2011 Adams Deaths 2462.00\n", "3 2011 Adams Group Quarters Population (July) 4024.00\n", "4 2011 Adams Household Population (July) 447552.00\n", "5 2011 Adams Housing Vacancy Rate 3.84\n", "6 2011 Adams July Population 451576.00\n", "7 2011 Adams Natural Population Increase 4782.00\n", "8 2011 Adams Net Building Permits 226.00\n", "9 2011 Adams Net Migration 3083.00\n", "10 2011 Adams Population (U.S. Census Bureau) 451443.00\n", "11 2011 Adams Total Households 157264.00\n", "12 2011 Adams Total Housing Units 163550.00\n", "13 2011 Adams Vacant Housing Units 6286.00\n", "14 2011 Alamosa Average Household Size 2.45\n", "15 2011 Alamosa Births 234.00\n", "16 2011 Alamosa Deaths 131.00\n", "17 2011 Alamosa Group Quarters Population (July) 740.00\n", "18 2011 Alamosa Household Population (July) 14902.00\n", "19 2011 Alamosa Housing Vacancy Rate 7.95\n", "20 2011 Alamosa July Population 15642.00\n", "21 2011 Alamosa Natural Population Increase 103.00\n", "22 2011 Alamosa Net Building Permits 8.00\n", "23 2011 Alamosa Net Migration 65.00\n", "24 2011 Alamosa Population (U.S. Census Bureau) 15710.00\n", "25 2011 Alamosa Total Households 6075.00\n", "26 2011 Alamosa Total Housing Units 6600.00\n", "27 2011 Alamosa Vacant Housing Units 525.00\n", "28 2011 Arapahoe Average Household Size 2.53\n", "29 2011 Arapahoe Births 7721.00\n", "30 2011 Arapahoe Deaths 3426.00\n", "31 2011 Arapahoe Group Quarters Population (July) 4894.00\n", "32 2011 Arapahoe Household Population (July) 579809.00\n", "33 2011 Arapahoe Housing Vacancy Rate 4.53\n", "34 2011 Arapahoe July Population 584703.00\n", "35 2011 Arapahoe Natural Population Increase 4295.00\n", "36 2011 Arapahoe Net Building Permits 1102.00\n", "37 2011 Arapahoe Net Migration 5589.00\n", "38 2011 Arapahoe Population (U.S. Census Bureau) 584948.00\n", "39 2011 Arapahoe Total Households 229014.00\n", "40 2011 Arapahoe Total Housing Units 239870.00\n", "41 2011 Arapahoe Vacant Housing Units 10856.00\n", "42 2011 Archuleta Average Household Size 2.27\n", "43 2011 Archuleta Births 137.00\n", "44 2011 Archuleta Deaths 82.00\n", "45 2011 Archuleta Group Quarters Population (July) 129.00\n", "46 2011 Archuleta Household Population (July) 11909.00\n", "47 2011 Archuleta Housing Vacancy Rate 40.44\n", "48 2011 Archuleta July Population 12038.00\n", "49 2011 Archuleta Natural Population Increase 55.00\n", "50 2011 Archuleta Net Building Permits 36.00\n", "51 2011 Archuleta Net Migration -77.00\n", "52 2011 Archuleta Population (U.S. Census Bureau) 12013.00\n", "53 2011 Archuleta Total Households 5247.00\n", "54 2011 Archuleta Total Housing Units 8810.00\n", "55 2011 Archuleta Vacant Housing Units 3563.00\n", "56 2011 Baca Average Household Size 2.20\n", "57 2011 Baca Births 41.00\n", "58 2011 Baca Deaths 73.00\n", "59 2011 Baca Group Quarters Population (July) 82.00\n", " ... ... ... ...\n", "\n", "[24192 rows x 4 columns]" ] } ], "prompt_number": 22 }, { "cell_type": "markdown", "metadata": {}, "source": [ "To integrate this set with the original data, it is best to unstack the variable column." ] }, { "cell_type": "code", "collapsed": false, "input": [ "#Temporarily set index to facilitate unstacking of 'var'\n", "co_d_temp=co_dem.set_index(['AUDIT_YEAR','FNAME','var']).unstack('var')\n", "\n", "#Rename columns (dropping value level from column names and abbreviating)\n", "desired_cols=['hh_size','births','deaths','gq_pop','hh_pop','vac_rate','pop','npop_increase',\\\n", " 'permits','n_migr','cens_pop','hh_num','hu_num','vac_num']\n", "co_d_temp.columns=desired_cols\n", "\n", "#Reset index\n", "co_d=co_d_temp.reset_index()\n", "\n", "print co_d\n", "\n", "co_d.head()" ], "language": "python", "metadata": {}, "outputs": [ { "output_type": "stream", "stream": "stdout", "text": [ " AUDIT_YEAR FNAME hh_size births deaths gq_pop hh_pop vac_rate pop npop_increase permits n_migr cens_pop hh_num \\\n", "0 1985 Adams 2.81 5210 1304 2256 266508 9.0 268764 3906 2741 -807 268764 94949 \n", "1 1985 Alamosa 2.84 258 79 884 12127 16.4 13011 179 10 -85 13011 4277 \n", "2 1985 Arapahoe 2.68 6458 1602 3536 377708 8.4 381244 4856 5069 9246 381244 140787 \n", "3 1985 Archuleta 2.84 101 22 0 5213 45.6 5213 79 9 149 5213 1837 \n", "4 1985 Baca 2.61 83 62 0 4873 27.7 4873 21 1 -195 4873 1867 \n", "5 1985 Bent 2.66 80 64 424 5275 21.6 5699 16 6 -155 5699 1986 \n", "6 1985 Boulder 2.54 3326 1064 8600 200844 10.3 209444 2262 1953 -3005 209444 79100 \n", "7 1985 Broomfield 0.00 0 0 0 0 0.0 0 0 0 0 0 0 \n", "8 1985 Chaffee 2.58 156 107 612 11737 29.0 12349 49 73 -538 12349 4543 \n", "9 1985 Cheyenne 2.36 55 21 14 2280 9.0 2294 34 3 23 2294 965 \n", "10 1985 Clear Creek 2.55 140 40 45 7891 32.7 7936 100 53 84 7936 3093 \n", "11 1985 Conejos 3.13 143 60 0 7776 25.0 7776 83 1 -312 7776 2486 \n", "12 1985 Costilla 2.88 50 20 0 3363 27.7 3363 30 0 73 3363 1167 \n", "13 1985 Crowley 2.33 51 36 53 2944 3.6 2997 15 1 94 2997 1262 \n", "14 1985 Custer 2.49 22 22 18 2048 36.7 2066 0 0 160 2066 823 \n", "15 1985 Delta 2.50 323 245 331 23135 7.4 23466 78 10 -617 23466 9237 \n", "16 1985 Denver 2.13 8952 4541 13031 483400 5.0 496431 4411 1272 -4595 496431 226540 \n", "17 1985 Dolores 2.71 27 9 0 1548 37.7 1548 18 0 -154 1548 571 \n", "18 1985 Douglas 3.37 679 129 163 39885 22.3 40048 550 2160 3214 40048 11824 \n", "19 1985 Eagle 2.40 386 41 6 18023 44.8 18029 345 74 923 18029 7498 \n", "20 1985 El Paso 2.63 7055 1851 17384 351122 10.6 368506 5204 6499 8645 368506 133562 \n", "21 1985 Elbert 2.93 109 42 35 8525 17.4 8560 67 184 233 8560 2908 \n", "22 1985 Fremont 2.61 364 374 2749 28617 16.3 31366 -10 259 93 31366 10955 \n", "23 1985 Garfield 2.68 467 172 444 25270 25.5 25714 295 56 -238 25714 9426 \n", "24 1985 Gilpin 2.69 41 17 51 3030 48.4 3081 24 36 31 3081 1128 \n", "25 1985 Grand 2.45 144 41 9 9203 62.3 9212 103 160 218 9212 3751 \n", "26 1985 Gunnison 2.21 140 50 1399 8991 43.2 10390 90 87 -117 10390 4072 \n", "27 1985 Hinsdale 2.81 5 0 0 472 81.1 472 5 20 -116 472 168 \n", "28 1985 Huerfano 2.40 85 81 144 6665 32.6 6809 4 38 165 6809 2778 \n", "29 1985 Jackson 2.85 28 7 19 1710 48.5 1729 21 7 -85 1729 600 \n", "30 1985 Jefferson 2.79 6675 2182 6206 415411 9.2 421617 4493 4909 3441 421617 148723 \n", "31 1985 Kiowa 2.50 24 12 29 1818 14.9 1847 12 0 -123 1847 728 \n", "32 1985 Kit Carson 2.55 119 69 70 7477 17.7 7547 50 2 -304 7547 2930 \n", "33 1985 La Plata 2.66 550 182 1243 30150 24.1 31393 368 304 -47 31393 11329 \n", "34 1985 Lake 2.73 109 33 96 6693 36.0 6789 76 8 -455 6789 2452 \n", "35 1985 Larimer 2.55 2785 962 8066 162658 12.2 170724 1823 2619 2437 170724 63873 \n", "36 1985 Las Animas 2.59 185 163 469 13772 20.0 14241 22 22 -496 14241 5313 \n", "37 1985 Lincoln 2.39 76 51 107 4481 18.5 4588 25 6 92 4588 1872 \n", "38 1985 Logan 2.50 288 163 624 18439 12.0 19063 125 12 -455 19063 7377 \n", "39 1985 Mesa 2.58 1423 682 2112 85909 18.8 88021 741 147 -6237 88021 33295 \n", "40 1985 Mineral 2.59 16 4 0 836 63.6 836 12 10 -40 836 323 \n", "41 1985 Moffat 2.89 256 63 97 13189 20.9 13286 193 3 -1305 13286 4567 \n", "42 1985 Montezuma 2.99 345 109 160 19123 15.9 19283 236 18 -784 19283 6390 \n", "43 1985 Montrose 2.62 348 238 441 23948 15.9 24389 110 108 -692 24389 9155 \n", "44 1985 Morgan 2.64 430 218 367 22064 12.9 22431 212 33 -594 22431 8353 \n", "45 1985 Otero 2.63 348 255 517 21243 12.2 21760 93 6 -172 21760 8064 \n", "46 1985 Ouray 2.74 33 10 0 2130 48.1 2130 23 13 -89 2130 776 \n", "47 1985 Park 3.02 101 31 14 6587 67.0 6601 70 228 -229 6601 2178 \n", "48 1985 Phillips 2.43 61 60 59 4362 15.3 4421 1 3 -10 4421 1794 \n", "49 1985 Pitkin 2.07 144 27 21 11295 42.1 11316 117 39 420 11316 5458 \n", "50 1985 Prowers 2.73 318 120 163 14211 14.6 14374 198 21 -126 14374 5205 \n", "51 1985 Pueblo 2.54 1844 1127 2613 119294 8.2 121907 717 272 -1205 121907 46877 \n", "52 1985 Rio Blanco 3.06 127 37 261 6329 31.4 6590 90 0 -441 6590 2071 \n", "53 1985 Rio Grande 2.85 240 115 279 11267 18.9 11546 125 0 68 11546 3959 \n", "54 1985 Routt 2.51 222 62 126 14070 38.6 14196 160 62 -216 14196 5596 \n", "55 1985 Saguache 3.14 73 34 17 4383 32.4 4400 39 2 -33 4400 1397 \n", "56 1985 San Juan 2.46 24 5 0 763 41.4 763 19 0 -174 763 310 \n", "57 1985 San Miguel 2.25 63 5 0 3189 32.7 3189 58 156 172 3189 1415 \n", "58 1985 Sedgwick 2.37 30 39 50 2941 18.4 2991 -9 0 -201 2991 1240 \n", "59 1985 Summit 2.26 243 25 0 12865 61.0 12865 218 115 633 12865 5697 \n", " ... ... ... ... ... ... ... ... ... ... ... ... ... ... \n", "\n", " hu_num vac_num \n", "0 104376 9427 \n", "1 5113 836 \n", "2 153726 12939 \n", "3 3379 1542 \n", "4 2581 714 \n", "5 2532 546 \n", "6 88178 9078 \n", "7 0 0 \n", "8 6402 1859 \n", "9 1060 95 \n", "10 4596 1503 \n", "11 3316 830 \n", "12 1614 447 \n", "13 1309 47 \n", "14 1300 477 \n", "15 9974 737 \n", "16 238346 11806 \n", "17 916 345 \n", "18 15225 3401 \n", "19 13585 6087 \n", "20 149466 15904 \n", "21 3520 612 \n", "22 13085 2130 \n", "23 12647 3221 \n", "24 2185 1057 \n", "25 9948 6197 \n", "26 7165 3093 \n", "27 888 720 \n", "28 4123 1345 \n", "29 1166 566 \n", "30 163850 15127 \n", "31 855 127 \n", "32 3560 630 \n", "33 14935 3606 \n", "34 3833 1381 \n", "35 72762 8889 \n", "36 6645 1332 \n", "37 2296 424 \n", "38 8381 1004 \n", "39 40995 7700 \n", "40 887 564 \n", "41 5772 1205 \n", "42 7601 1211 \n", "43 10891 1736 \n", "44 9588 1235 \n", "45 9181 1117 \n", "46 1496 720 \n", "47 6600 4422 \n", "48 2118 324 \n", "49 9432 3974 \n", "50 6097 892 \n", "51 51045 4168 \n", "52 3019 948 \n", "53 4879 920 \n", "54 9111 3515 \n", "55 2067 670 \n", "56 529 219 \n", "57 2103 688 \n", "58 1519 279 \n", "59 14623 8926 \n", " ... ... \n", "\n", "[1728 rows x 16 columns]\n" ] }, { "html": [ "

	AUDIT_YEAR	FNAME	hh_size	births	deaths	gq_pop	hh_pop	vac_rate	pop	npop_increase	permits	n_migr	cens_pop	hh_num	hu_num	vac_num
0	1985	Adams	2.81	5210	1304	2256	266508	9.0	268764	3906	2741	-807	268764	94949	104376	9427
1	1985	Alamosa	2.84	258	79	884	12127	16.4	13011	179	10	-85	13011	4277	5113	836
2	1985	Arapahoe	2.68	6458	1602	3536	377708	8.4	381244	4856	5069	9246	381244	140787	153726	12939
3	1985	Archuleta	2.84	101	22	0	5213	45.6	5213	79	9	149	5213	1837	3379	1542
4	1985	Baca	2.61	83	62	0	4873	27.7	4873	21	1	-195	4873	1867	2581	714

\n", "

5 rows \u00d7 16 columns

\n", "

" ], "metadata": {}, "output_type": "pyout", "prompt_number": 23, "text": [ " AUDIT_YEAR FNAME hh_size births deaths gq_pop hh_pop vac_rate pop npop_increase permits n_migr cens_pop hh_num \\\n", "0 1985 Adams 2.81 5210 1304 2256 266508 9.0 268764 3906 2741 -807 268764 94949 \n", "1 1985 Alamosa 2.84 258 79 884 12127 16.4 13011 179 10 -85 13011 4277 \n", "2 1985 Arapahoe 2.68 6458 1602 3536 377708 8.4 381244 4856 5069 9246 381244 140787 \n", "3 1985 Archuleta 2.84 101 22 0 5213 45.6 5213 79 9 149 5213 1837 \n", "4 1985 Baca 2.61 83 62 0 4873 27.7 4873 21 1 -195 4873 1867 \n", "\n", " hu_num vac_num \n", "0 104376 9427 \n", "1 5113 836 \n", "2 153726 12939 \n", "3 3379 1542 \n", "4 2581 714 \n", "\n", "[5 rows x 16 columns]" ] } ], "prompt_number": 23 }, { "cell_type": "markdown", "metadata": {}, "source": [ "From here we can merge these sets on ['AUDIT_YEAR','FNAME']..." ] }, { "cell_type": "code", "collapsed": false, "input": [ "co_sfd=pd.merge(co_sf,co_d,on=['AUDIT_YEAR','FNAME'])\n", "\n", "print co_sfd\n", "\n", "co_sfd.columns" ], "language": "python", "metadata": {}, "outputs": [ { "output_type": "stream", "stream": "stdout", "text": [ " Unnamed: 0 ALAND AWATER CBSAFP CLASSFP COUNTYFP COUNTYNS CSAFP FUNCSTAT GEOID INTPTLAT INTPTLON LSAD METDIVFP \\\n", "0 0 2893573835 15324680 NaN H1 53 198142 NaN A 8053 37.811667 -107.383335 6 NaN \n", "1 1 2957922894 16953424 NaN H1 29 198130 NaN A 8029 38.861756 -107.864757 6 NaN \n", "2 2 388229141 941898 19740 H1 47 198139 216 A 8047 39.861082 -105.528947 6 NaN \n", "3 3 3177803460 8828904 NaN H1 23 198127 NaN A 8023 37.277547 -105.428940 6 NaN \n", "4 4 4179563675 18750018 NaN H1 57 198144 NaN A 8057 40.663432 -106.329248 6 NaN \n", "5 5 2624702134 4013482 NaN H1 15 198123 NaN A 8015 38.738244 -106.316955 6 NaN \n", "6 6 4793688729 442148 19740 H1 39 198136 216 A 8039 39.310582 -104.117729 6 NaN \n", "7 7 1442738671 4944573 17820 H1 119 198175 NaN A 8119 38.871994 -105.182552 6 NaN \n", "8 8 4243419913 15352633 NaN H1 99 198165 NaN A 8099 37.958181 -102.392161 6 NaN \n", "9 9 1419419287 3530746 NaN H1 115 198173 NaN A 8115 40.871568 -102.355358 6 NaN \n", "10 10 396198354 4222992 19740 H6 31 198131 216 C 8031 39.761850 -104.880641 6 NaN \n", "11 11 2361929110 984917 NaN H1 105 198168 NaN A 8105 37.485763 -106.453214 6 NaN \n", "12 12 5256407750 27516850 NaN H1 83 198157 NaN A 8083 37.338025 -108.595786 6 NaN \n", "13 13 12284921637 19806932 NaN H1 81 198156 NaN A 8081 40.573984 -108.204521 6 NaN \n", "14 14 3334263504 9235291 NaN H1 21 198126 NaN A 8021 37.213406 -106.176447 6 NaN \n", "15 15 4382658416 19545450 20420 H1 67 198148 NaN A 8067 37.287367 -107.839718 6 NaN \n", "16 16 2763645279 2716141 NaN H1 33 198132 NaN A 8033 37.747602 -108.530383 6 NaN \n", "17 17 5596501541 2251930 NaN H1 63 198147 NaN A 8063 39.305340 -102.603023 6 NaN \n", "18 18 10326885138 76675362 24540 H1 123 198177 216 A 8123 40.555794 -104.383649 6 NaN \n", "19 19 4578600049 47032937 NaN H1 61 198146 NaN A 8061 38.388466 -102.756210 6 NaN \n", "20 20 8207226426 4454545 NaN H1 109 198170 NaN A 8109 38.031651 -106.234666 6 NaN \n", "21 21 6521817532 15207414 NaN H1 121 198176 NaN A 8121 39.965790 -103.209744 6 NaN \n", "22 22 3496904695 13827548 NaN H1 7 198119 NaN A 8007 37.202395 -107.050863 6 NaN \n", "23 23 6617316468 6142193 NaN H1 9 198120 NaN A 8009 37.306678 -102.537765 6 NaN \n", "24 24 2176207910 6769100 19740 H1 35 198133 216 A 8035 39.326435 -104.926199 6 NaN \n", "25 25 6179976224 30284066 39380 H1 101 198166 NaN A 8101 38.170658 -104.489893 6 NaN \n", "26 26 4362916056 18803966 20780 H1 37 198134 NaN A 8037 39.630638 -106.692944 6 NaN \n", "27 27 1781724999 301808 NaN H1 95 198163 NaN A 8095 40.594712 -102.345105 6 NaN \n", "28 28 8342198615 4868111 NaN H1 103 198167 NaN A 8103 39.972606 -108.200685 6 NaN \n", "29 29 6117633825 15831696 NaN H1 107 198169 NaN A 8107 40.483155 -106.985274 6 NaN \n", "30 30 1575562624 28287714 43540 H1 117 198174 NaN A 8117 39.621023 -106.137554 6 NaN \n", "31 31 3024201147 41944233 19740 H1 1 198116 216 A 8001 39.874325 -104.331872 6 NaN \n", "32 32 4762050035 16363552 44540 H1 75 198153 NaN A 8075 40.728091 -103.090464 6 NaN \n", "33 33 5803398598 4923261 33940 H1 85 198158 NaN A 8085 38.413427 -108.263042 6 NaN \n", "34 34 7634471698 21430626 NaN H1 45 198138 NaN A 8045 39.599352 -107.909780 6 NaN \n", "35 35 5508460909 7111548 17820 H1 41 198135 NaN A 8041 38.827383 -104.527472 6 NaN \n", "36 36 2268230027 5222170 NaN H1 79 198155 NaN A 8079 37.651478 -106.932300 6 NaN \n", "37 37 5682134218 43530467 19740 H1 93 198162 216 A 8093 39.118914 -105.717648 6 NaN \n", "38 38 4120753930 5792107 NaN H1 55 198143 NaN A 8055 37.687815 -104.959928 6 NaN \n", "39 39 3918307819 73093700 NaN H1 11 198121 NaN A 8011 37.931891 -103.077584 6 NaN \n", "40 40 1023554447 3280081 19740 H1 19 198125 216 A 8019 39.689402 -105.670791 6 NaN \n", "41 41 976206766 18110615 20780 H1 65 198149 NaN A 8065 39.205341 -106.350079 6 NaN \n", "42 42 4605714106 8166145 NaN H1 17 198124 NaN A 8017 38.835386 -102.604585 6 NaN \n", "43 43 12361253683 6929976 NaN H1 71 198151 NaN A 8071 37.318831 -104.044110 6 NaN \n", "44 44 1913033457 3364740 NaN H1 27 198129 NaN A 8027 38.101994 -105.373515 6 NaN \n", "45 45 1871620246 1847609 NaN H1 3 198117 NaN A 8003 37.568442 -105.788041 6 NaN \n", "46 46 6123751576 11134682 NaN H1 125 198178 NaN A 8125 40.000631 -102.422649 6 NaN \n", "47 47 8622131029 31435336 24300 H1 77 198154 NaN A 8077 39.019492 -108.461837 6 NaN \n", "48 48 4781807818 60250181 NaN H1 49 198140 NaN A 8049 40.123289 -106.095876 6 NaN \n", "49 49 1402725842 1599543 NaN H1 91 198161 NaN A 8091 38.150600 -107.767133 6 NaN \n", "50 50 8389491951 53067063 NaN H1 51 198141 NaN A 8051 38.669679 -107.078108 6 NaN \n", "51 51 3332207690 5213860 NaN H1 113 198172 NaN A 8113 38.009373 -108.427326 6 NaN \n", "52 52 3316112929 34663780 22820 H1 87 198159 NaN A 8087 40.262540 -103.807497 6 NaN \n", "53 53 6723779313 98113981 22660 H1 69 198150 NaN A 8069 40.663091 -105.482131 6 NaN \n", "54 54 3268485084 20052353 NaN H1 89 198160 NaN A 8089 37.884170 -103.721260 6 NaN \n", "55 55 2514071663 6469378 NaN H1 97 198164 NaN A 8097 39.217533 -106.915943 6 NaN \n", "56 56 6676208280 22780444 NaN H1 73 198152 NaN A 8073 38.993740 -103.507554 6 NaN \n", "57 57 2039409780 33430387 NaN H1 25 198128 NaN A 8025 38.306180 -103.772736 6 NaN \n", "58 58 1003640124 2035929 NaN H1 111 198171 NaN A 8111 37.781075 -107.670257 6 NaN \n", "59 59 3970604102 2235391 15860 H1 43 198137 NaN A 8043 38.455658 -105.421438 6 NaN \n", " ... ... ... ... ... ... ... ... ... ... ... ... ... ... \n", "\n", " MTFCC NAME NAMELSAD STATEFP sort_id index \n", "0 G4020 Hinsdale Hinsdale County 8 0 902 ... \n", "1 G4020 Delta Delta County 8 1 626 ... \n", "2 G4020 Gilpin Gilpin County 8 2 833 ... \n", "3 G4020 Costilla Costilla County 8 3 516 ... \n", "4 G4020 Jackson Jackson County 8 4 948 ... \n", "5 G4020 Chaffee Chaffee County 8 5 332 ... \n", "6 G4020 Elbert Elbert County 8 6 741 ... \n", "7 G4020 Teller Teller County 8 7 425 ... \n", "8 G4020 Prowers Prowers County 8 8 1431 ... \n", "9 G4020 Sedgwick Sedgwick County 8 9 333 ... \n", "10 G4020 Denver Denver County 8 10 649 ... \n", "11 G4020 Rio Grande Rio Grande County 8 11 103 ... \n", "12 G4020 Montezuma Montezuma County 8 12 1247 ... \n", "13 G4020 Moffat Moffat County 8 13 1224 ... \n", "14 G4020 Conejos Conejos County 8 14 470 ... \n", "15 G4020 La Plata La Plata County 8 15 1063 ... \n", "16 G4020 Dolores Dolores County 8 16 672 ... \n", "17 G4020 Kit Carson Kit Carson County 8 17 1017 ... \n", "18 G4020 Weld Weld County 8 18 517 ... \n", "19 G4020 Kiowa Kiowa County 8 19 994 ... \n", "20 G4020 Saguache Saguache County 8 20 195 ... \n", "21 G4020 Washington Washington County 8 21 471 ... \n", "22 G4020 Archuleta Archuleta County 8 22 148 ... \n", "23 G4020 Baca Baca County 8 23 194 ... \n", "24 G4020 Douglas Douglas County 8 24 695 ... \n", "25 G4020 Pueblo Pueblo County 8 25 11 ... \n", "26 G4020 Eagle Eagle County 8 26 718 ... \n", "27 G4020 Phillips Phillips County 8 27 1385 ... \n", "28 G4020 Rio Blanco Rio Blanco County 8 28 57 ... \n", "29 G4020 Routt Routt County 8 29 149 ... \n", "30 G4020 Summit Summit County 8 30 379 ... \n", "31 G4020 Adams Adams County 8 31 10 ... \n", "32 G4020 Logan Logan County 8 32 1155 ... \n", "33 G4020 Montrose Montrose County 8 33 1270 ... \n", "34 G4020 Garfield Garfield County 8 34 810 ... \n", "35 G4020 El Paso El Paso County 8 35 764 ... \n", "36 G4020 Mineral Mineral County 8 36 1201 ... \n", "37 G4020 Park Park County 8 37 1362 ... \n", "38 G4020 Huerfano Huerfano County 8 38 925 ... \n", "39 G4020 Bent Bent County 8 39 240 ... \n", "40 G4020 Clear Creek Clear Creek County 8 40 424 ... \n", "41 G4020 Lake Lake County 8 41 1040 ... \n", "42 G4020 Cheyenne Cheyenne County 8 42 378 ... \n", "43 G4020 Las Animas Las Animas County 8 43 1109 ... \n", "44 G4020 Custer Custer County 8 44 603 ... \n", "45 G4020 Alamosa Alamosa County 8 45 56 ... \n", "46 G4020 Yuma Yuma County 8 46 563 ... \n", "47 G4020 Mesa Mesa County 8 47 1178 ... \n", "48 G4020 Grand Grand County 8 48 856 ... \n", "49 G4020 Ouray Ouray County 8 49 1339 ... \n", "50 G4020 Gunnison Gunnison County 8 50 879 ... \n", "51 G4020 San Miguel San Miguel County 8 51 287 ... \n", "52 G4020 Morgan Morgan County 8 52 1293 ... \n", "53 G4020 Larimer Larimer County 8 53 1086 ... \n", "54 G4020 Otero Otero County 8 54 1316 ... \n", "55 G4020 Pitkin Pitkin County 8 55 1408 ... \n", "56 G4020 Lincoln Lincoln County 8 56 1132 ... \n", "57 G4020 Crowley Crowley County 8 57 562 ... \n", "58 G4020 San Juan San Juan County 8 58 241 ... \n", "59 G4020 Fremont Fremont County 8 59 787 ... \n", " ... ... ... ... ... ... \n", "\n", "[1472 rows x 138 columns]\n" ] }, { "metadata": {}, "output_type": "pyout", "prompt_number": 24, "text": [ "Index([u'Unnamed: 0', u'ALAND', u'AWATER', u'CBSAFP', u'CLASSFP', u'COUNTYFP', u'COUNTYNS', u'CSAFP', u'FUNCSTAT', u'GEOID', u'INTPTLAT', u'INTPTLON', u'LSAD', u'METDIVFP', u'MTFCC', u'NAME', u'NAMELSAD', u'STATEFP', u'sort_id', u'index', u'FNAME', u'AUDIT_YEAR', u'REV_TOTAL', u'REV_INTGOVT', u'EXP_TOTAL', u'TOTAL_DEBT', u'CTY_POP', u'intensity', u'pcrev', u'pcexp', u'pcintgov', u'pcdebt', u'pc_ap', u'w_rook_i', u'w_queen_i', u'w_db_b_i', u'w_db_c_i', u'w_kern_i', u'w_rook_i_ap', u'w_queen_i_ap', u'w_db_b_i_ap', u'w_db_c_i_ap', u'w_kern_i_ap', u'w_rook_g', u'w_queen_g', u'w_db_b_g', u'w_db_c_g', u'w_kern_g', u'w_rook_g_ap', u'w_queen_g_ap', u'w_db_b_g_ap', u'w_db_c_g_ap', u'w_kern_g_ap', u'fipsc', u'fipscty', u'year', u'ap', u'pcap', u'pcap_mean', u'pcap_rel', u'gsp', u'pcap_rel2', u'LG_ID', u'REV_CTF', u'EXP_DEBT_SERVICE_GEN', u'EXP_TRANSFER_OUT', u'ST_SALES_TAX_PAID', u'EXP_CAPITAL_OUTLAY', u'EXP_JUDICIAL', u'db', u'EXP_HEALTH', u'EXP_FIRE', u'EXP_OTHER_PUBLIC_WORKS', u'REV_FRANCHISE_TAX', u'POPULATION', u'EXP_GEN_GOVT', u'LIABILITIES', u'EXP_TOTAL_PUBLIC_SAFETY', u'REV_LODGING_TAX', u'ALL_OTHER_INTGOVT', u'REVENUE_DEBT_GEN', u'EXP_MISC', u'REV_OTHER_TAX', u'REV_UNCLASS_TAX', u'db_temp', u'GO_DEBT_GEN', u'OTHER_DEBT_GEN', u'sorting_id', u'EXP_SOCIAL_SERVICE', u'EXP_PRINCIPAL_GEN', u'REV_MOTOR_VEH_FEE', u'REV_TRANSFER_IN', u'EXP_RECREATION', u'EXP_POLICE', u'EXP_INTEREST_GEN', u'NAME_x', u'NAME_y', u'EXP_OTHER_PUBLIC_SAFETY', u'REV_CHARGES', u'EXP_TOTAL_PUBLIC_WORKS', ...], dtype='object')" ] } ], "prompt_number": 24 }, { "cell_type": "markdown", "metadata": {}, "source": [ "Moving on to assessment data, we can leverage the file provided by the Colorado Division of Local Affairs that provides time series data on assessment base by county and class of property. The data were summarized into residential and non-residential categories in the script called *tabor_score*. Further, the market values were estimated by leveraging residential and non-residential rate history. *This is an imperfect measure insofar as applying the rate to these group portions of the base ignores both exempt property value and the differential treatment of personal property*. The former issue is not a big deal, because we really only care about the growth in the taxable base. The latter, however, is an issue since we do not know how much of the base personal property commands. \n", "\n", "I should ask whether personal property is housed solely under the Residential class, or if it occurs elsewhere as well.\n", "\n", "In any event, here is the summary data..." ] }, { "cell_type": "code", "collapsed": false, "input": [ "co_asmt=pd.read_csv(workdir+'co_cty_asmt.csv')[pd.read_csv(workdir+'co_cty_asmt.csv').columns[1:]]\n", "co_asmt['cty'].replace('Park\\\\','Park',inplace=True)\n", "\n", "print co_asmt\n", "\n", "HTML(co_asmt.head(10).to_html())" ], "language": "python", "metadata": {}, "outputs": [ { "output_type": "stream", "stream": "stdout", "text": [ " cty resid total year non_resid res_rate nonres_rate res_val nonres_val\n", "0 Adams 705269720 1632609350 1993 927339630 0.1286 0.29 5.484212e+09 3.197723e+09\n", "1 Adams 742865190 1684395240 1994 941530050 0.1286 0.29 5.776557e+09 3.246655e+09\n", "2 Adams 813385250 1814600930 1995 1001215680 0.1036 0.29 7.851209e+09 3.452468e+09\n", "3 Adams 853089120 1898365960 1996 1045276840 0.1036 0.29 8.234451e+09 3.604403e+09\n", "4 Adams 955203880 2094081960 1997 1138878080 0.0974 0.29 9.807021e+09 3.927166e+09\n", "5 Adams 1003073550 2204397180 1998 1201323630 0.0974 0.29 1.029850e+10 4.142495e+09\n", "6 Adams 1181599650 2552946760 1999 1371347110 0.0974 0.29 1.213141e+10 4.728783e+09\n", "7 Adams 1257611690 2706528490 2000 1448916800 0.0974 0.29 1.291182e+10 4.996265e+09\n", "8 Adams 1568125310 3301114460 2001 1732989150 0.0915 0.29 1.713798e+10 5.975825e+09\n", "9 Adams 1581702560 3343110670 2002 1761408110 0.0915 0.29 1.728637e+10 6.073821e+09\n", "10 Adams 1730702090 3609794180 2003 1879092090 0.0796 0.29 2.174249e+10 6.479628e+09\n", "11 Adams 1824215350 3765774100 2004 1941558750 0.0796 0.29 2.291728e+10 6.695030e+09\n", "12 Adams 2005601680 4118305220 2005 2112703540 0.0796 0.29 2.519600e+10 7.285185e+09\n", "13 Adams 2090024670 4246997040 2006 2156972370 0.0796 0.29 2.625659e+10 7.437836e+09\n", "14 Adams 2136421180 4524060000 2007 2387638820 0.0796 0.29 2.683946e+10 8.233237e+09\n", "15 Adams 2177583150 4659254010 2008 2481670860 0.0796 0.29 2.735657e+10 8.557486e+09\n", "16 Adams 2005689060 4598155740 2009 2592466680 0.0796 0.29 2.519710e+10 8.939540e+09\n", "17 Adams 2015295810 4601619680 2010 2586323870 0.0796 0.29 2.531779e+10 8.918358e+09\n", "18 Adams 1966947450 4568563790 2011 2601616340 0.0796 0.29 2.471040e+10 8.971091e+09\n", "19 Adams 1983416840 4622808830 2012 2639391990 0.0796 0.29 2.491730e+10 9.101352e+09\n", "20 Alamosa 20978680 64233640 1993 43254960 0.1286 0.29 1.631313e+08 1.491550e+08\n", "21 Alamosa 21551860 68170650 1994 46618790 0.1286 0.29 1.675883e+08 1.607544e+08\n", "22 Alamosa 21090890 68533240 1995 47442350 0.1036 0.29 2.035800e+08 1.635943e+08\n", "23 Alamosa 22275310 70148290 1996 47872980 0.1036 0.29 2.150126e+08 1.650792e+08\n", "24 Alamosa 25394010 78460170 1997 53066160 0.0974 0.29 2.607188e+08 1.829868e+08\n", "25 Alamosa 26214120 81288350 1998 55074230 0.0974 0.29 2.691388e+08 1.899111e+08\n", "26 Alamosa 29715900 88056760 1999 58340860 0.0974 0.29 3.050914e+08 2.011754e+08\n", "27 Alamosa 30653970 92072460 2000 61418490 0.0974 0.29 3.147225e+08 2.117879e+08\n", "28 Alamosa 33900440 99209940 2001 65309500 0.0915 0.29 3.704966e+08 2.252052e+08\n", "29 Alamosa 34958800 105082050 2002 70123250 0.0915 0.29 3.820634e+08 2.418043e+08\n", "30 Alamosa 32428230 105710210 2003 73281980 0.0796 0.29 4.073898e+08 2.526965e+08\n", "31 Alamosa 33207320 105759120 2004 72551800 0.0796 0.29 4.171774e+08 2.501786e+08\n", "32 Alamosa 35231910 113201810 2005 77969900 0.0796 0.29 4.426119e+08 2.688617e+08\n", "33 Alamosa 36344730 115502520 2006 79157790 0.0796 0.29 4.565921e+08 2.729579e+08\n", "34 Alamosa 43989700 128487250 2007 84497550 0.0796 0.29 5.526344e+08 2.913709e+08\n", "35 Alamosa 45135218 131937947 2008 86802729 0.0796 0.29 5.670254e+08 2.993198e+08\n", "36 Alamosa 49620684 144285542 2009 94664858 0.0796 0.29 6.233754e+08 3.264305e+08\n", "37 Alamosa 49791969 142978579 2010 93186610 0.0796 0.29 6.255272e+08 3.213331e+08\n", "38 Alamosa 50715636 147822295 2011 97106659 0.0796 0.29 6.371311e+08 3.348505e+08\n", "39 Alamosa 51319469 150635429 2012 99315960 0.0796 0.29 6.447169e+08 3.424688e+08\n", "40 Arapahoe 1716659900 3132892890 1993 1416232990 0.1286 0.29 1.334883e+10 4.883562e+09\n", "41 Arapahoe 1775102720 3172018980 1994 1396916260 0.1286 0.29 1.380329e+10 4.816953e+09\n", "42 Arapahoe 1909049660 3431555260 1995 1522505600 0.1036 0.29 1.842712e+10 5.250019e+09\n", "43 Arapahoe 1959244590 3572313810 1996 1613069220 0.1036 0.29 1.891163e+10 5.562308e+09\n", "44 Arapahoe 2111998320 4154999890 1997 2043001570 0.0974 0.29 2.168376e+10 7.044833e+09\n", "45 Arapahoe 2171015020 4299477620 1998 2128462600 0.0974 0.29 2.228968e+10 7.339526e+09\n", "46 Arapahoe 2498090430 5339974370 1999 2841883940 0.0974 0.29 2.564775e+10 9.799600e+09\n", "47 Arapahoe 2592694970 5563291240 2000 2970596270 0.0974 0.29 2.661904e+10 1.024344e+10\n", "48 Arapahoe 3266624560 6603331570 2001 3336707010 0.0915 0.29 3.570081e+10 1.150589e+10\n", "49 Arapahoe 3391718970 6710546210 2002 3318827240 0.0915 0.29 3.706797e+10 1.144423e+10\n", "50 Arapahoe 3442883070 6586625570 2003 3143742500 0.0796 0.29 4.325230e+10 1.084049e+10\n", "51 Arapahoe 3525994730 6655682660 2004 3129687930 0.0796 0.29 4.429642e+10 1.079203e+10\n", "52 Arapahoe 3694072400 6792660190 2005 3098587790 0.0796 0.29 4.640794e+10 1.068479e+10\n", "53 Arapahoe 3793975220 6921051070 2006 3127075850 0.0796 0.29 4.766301e+10 1.078302e+10\n", "54 Arapahoe 4150158740 7686129630 2007 3535970890 0.0796 0.29 5.213767e+10 1.219300e+10\n", "55 Arapahoe 4233898200 7798433580 2008 3564535380 0.0796 0.29 5.318968e+10 1.229150e+10\n", "56 Arapahoe 3955280450 7959760080 2009 4004479630 0.0796 0.29 4.968945e+10 1.380855e+10\n", "57 Arapahoe 3988026170 7963447430 2010 3975421260 0.0796 0.29 5.010083e+10 1.370835e+10\n", "58 Arapahoe 3774652610 7428001820 2011 3653349210 0.0796 0.29 4.742026e+10 1.259776e+10\n", "59 Arapahoe 3794965820 7461738443 2012 3666772623 0.0796 0.29 4.767545e+10 1.264404e+10\n", " ... ... ... ... ... ... ... ... ...\n", "\n", "[1280 rows x 9 columns]\n" ] }, { "html": [ "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "

	cty	resid	total	year	non_resid	res_rate	nonres_rate	res_val	nonres_val
0	Adams	705269720	1632609350	1993	927339630	0.1286	0.29	5.484212e+09	3.197723e+09
1	Adams	742865190	1684395240	1994	941530050	0.1286	0.29	5.776557e+09	3.246655e+09
2	Adams	813385250	1814600930	1995	1001215680	0.1036	0.29	7.851209e+09	3.452468e+09
3	Adams	853089120	1898365960	1996	1045276840	0.1036	0.29	8.234451e+09	3.604403e+09
4	Adams	955203880	2094081960	1997	1138878080	0.0974	0.29	9.807021e+09	3.927166e+09
5	Adams	1003073550	2204397180	1998	1201323630	0.0974	0.29	1.029850e+10	4.142495e+09
6	Adams	1181599650	2552946760	1999	1371347110	0.0974	0.29	1.213141e+10	4.728783e+09
7	Adams	1257611690	2706528490	2000	1448916800	0.0974	0.29	1.291182e+10	4.996265e+09
8	Adams	1568125310	3301114460	2001	1732989150	0.0915	0.29	1.713798e+10	5.975825e+09
9	Adams	1581702560	3343110670	2002	1761408110	0.0915	0.29	1.728637e+10	6.073821e+09

" ], "metadata": {}, "output_type": "pyout", "prompt_number": 25, "text": [ "" ] } ], "prompt_number": 25 }, { "cell_type": "markdown", "metadata": {}, "source": [ "These data must be deflated using the same CPI figures employed in deflating the fiscal data. They come from the [FRED](http://research.stlouisfed.org/fred2/) database." ] }, { "cell_type": "code", "collapsed": false, "input": [ "#Open feed to CPI info\n", "#cpi=web.get_data_fred('USACPIBLS','1/1/1975','1/1/2009')\n", "\n", "#Create copy in case offline work is occurring\n", "#cpi.to_csv(workdir+'us_cpi_1975_2009.csv')\n", "\n", "#Read in hard copy when API queries are exhausted\n", "cpi=pd.read_csv(workdir+'us_cpi_1975_2009.csv')\n", "\n", "#Create common index for later join with data\n", "cpi['AUDIT_YEAR']=range(1975,2010)\n", "cpi2=cpi.reset_index().set_index('AUDIT_YEAR')\n", "\n", "#Rename columns\n", "cpi2.columns=['index','DATE','CPI']\n", "\n", "#Drop DATE and index\n", "cpi2.pop('DATE')\n", "cpi2.pop('index')\n", "\n", "#Calculate deflator ratios\n", "dfl=cpi2.div(cpi2.ix[2009])\n", "\n", "#Reintegrate initial CPI figures\n", "dfl['cpi']=cpi2['CPI']\n", "\n", "#Calculate annual growth in inflation\n", "dfl['cpi_growth']=dfl['cpi'].pct_change()\n", "\n", "#Rename columns\n", "dfl.columns=['defl','cpi','cpi_growth']\n", "\n", "dfl.ix[1998:].head()" ], "language": "python", "metadata": {}, "outputs": [ { "html": [ "

	defl	cpi	cpi_growth
AUDIT_YEAR
1998	0.759907	163.0	0.015576
1999	0.776690	166.6	0.022086
2000	0.802797	172.2	0.033613
2001	0.825641	177.1	0.028455
2002	0.838695	179.9	0.015810

\n", "

5 rows \u00d7 3 columns

\n", "

" ], "metadata": {}, "output_type": "pyout", "prompt_number": 26, "text": [ " defl cpi cpi_growth\n", "AUDIT_YEAR \n", "1998 0.759907 163.0 0.015576\n", "1999 0.776690 166.6 0.022086\n", "2000 0.802797 172.2 0.033613\n", "2001 0.825641 177.1 0.028455\n", "2002 0.838695 179.9 0.015810\n", "\n", "[5 rows x 3 columns]" ] } ], "prompt_number": 26 }, { "cell_type": "markdown", "metadata": {}, "source": [ "Now the assessment data must actually be delated." ] }, { "cell_type": "code", "collapsed": false, "input": [ "#Merge deflator info in with assessment data\n", "co_asmt_temp=pd.merge(co_asmt,dfl.reset_index(),how='left',left_on='year',right_on='AUDIT_YEAR')\n", "\n", "#Set index\n", "co_asmt2=co_asmt_temp.set_index(['year','cty'])\n", "\n", "#Divide the level columns ['resid','total','non_resid','res_val','nonres_va'] by the deflator\n", "co_asmt_real=co_asmt2[['resid','total','non_resid','res_val','nonres_val']].div(co_asmt2['defl'],axis=0)\n", "\n", "#Rename columns\n", "co_asmt_real.columns=['real_resid','real_total','real_non_resid','real_res_val','real_nonres_val']\n", "\n", "#Merge real values back in\n", "co_asmt_r=pd.merge(co_asmt_real,co_asmt2,left_index=True,right_index=True)" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 27 }, { "cell_type": "markdown", "metadata": {}, "source": [ "These data must also be integrated into the master set." ] }, { "cell_type": "code", "collapsed": false, "input": [ "co_sfda=pd.merge(co_sfd,co_asmt_r.reset_index(),how='left',left_on=['AUDIT_YEAR','FNAME'], \\\n", " right_on=['year','cty']).rename(columns={'AUDIT_YEAR_x':'AUDIT_YEAR'})\n", "\n", "# for i in range(30):\n", "# print co_sfda[co_sfda.columns[i*5:(i+1)*5]].head(),'\\n'\n", " \n", "# HTML(co_sfda.describe().T.to_html())\n", "\n", "#Write to disk\n", "co_sfda.to_csv(workdir+'CO_sfda.csv')" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 28 }, { "cell_type": "markdown", "metadata": {}, "source": [ "Finally, we need business cycle information by county. The county-specific unemployment rate data (1990-2012) provided by the [Bureau of Labor Statistics](http://www.bls.gov/lau/#cntyaa)." ] }, { "cell_type": "code", "collapsed": false, "input": [ "# def labor_data(year):\n", "# '''Reads BLS data from web source as fixed width, converts to DataFrame, and subset to Colorado'''\n", "# #Set column widths (from 2012 example)\n", "# column_width_list=[(0,9),(11,15),(18,22),(24,72),(73,78),(83,93),(95,105),(108,117),(121,126)]\n", " \n", "# #Read data from BLS website\n", "# ue=pd.read_fwf('http://www.bls.gov/lau/laucnty'+year+'.txt',colspecs=column_width_list,skiprows=5)\n", " \n", "# #Rename columns\n", "# ue.columns=['laus_code','st_fips','cty_fips','cty_name','AUDIT_YEAR','labforce','emp','unemp','unemp_rate']\n", "\n", "# #Subset to Colorado\n", "# if isinstance(ue['st_fips'],basestring):\n", "# ue_co=ue[ue['st_fips']=='08']\n", "# else:\n", "# ue_co=ue[ue['st_fips']==8]\n", "# print ue.head()\n", "# print ue_co.head() \n", " \n", "# return ue_co\n", "\n", "# #Create containers for year specific data\n", "# labor_df_list=[]\n", "\n", "# #For each year...\n", "# for i in range(1990,2013):\n", "# print i\n", "# #...read and process data, and then throw it in the container\n", "# labor_df_list.append(labor_data(str(i)[2:]))\n", "\n", "# #Concatenate to one DF\n", "# co_labor=pd.concat(labor_df_list)\n", "\n", "# print co_labor" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 29 }, { "cell_type": "markdown", "metadata": {}, "source": [ "In tinkering with the function, it looks like I have exhausted my BLS API calls for the day, so I just downloaded the table into an intermediate .csv. Furthermore, rather than read and convert state unemplyment rates, it was easier just to grab the data via a dictionary." ] }, { "cell_type": "code", "collapsed": false, "input": [ "#Read in data, and stack year values\n", "co_bls=pd.read_csv(workdir+'CO_BLS_labor.csv').set_index('Series_ID').stack().reset_index()\n", "\n", "#Rename columns\n", "co_bls.columns=['Series_ID','AUDIT_YEAR_temp','val']\n", "\n", "#Convert AUDIT_YEAR to int and drop temp\n", "co_bls['AUDIT_YEAR']=co_bls['AUDIT_YEAR_temp'].astype(int)\n", "co_bls.pop('AUDIT_YEAR_temp')\n", "\n", "#Split out series indicator component of the Series ID\n", "co_bls['indic_code']=co_bls['Series_ID'].apply(lambda x: x[-3:])\n", "\n", "#Create dict to map descriptions to indic_code\n", "indic_dict={'003':'unemp_rate',\n", " '004':'unemp',\n", " '005':'emp',\n", " '006':'labforce'}\n", "\n", "#Define new indicator description variable\n", "co_bls['indic']=co_bls['indic_code'].map(indic_dict)\n", "\n", "#Split out county fips ID and area type from Series_ID\n", "co_bls['fipscty_str']=co_bls['Series_ID'].apply(lambda x: x[7:10])\n", "co_bls['fipscty']=co_bls['fipscty_str'].astype(int)\n", "co_bls['atype']=co_bls['Series_ID'].apply(lambda x: x[3:5])\n", "\n", "#Drop the now useless Series_ID and indic_code\n", "co_bls.pop('Series_ID')\n", "co_bls.pop('indic_code')\n", "\n", "#Unstack indicators into separate columns\n", "co_labor_temp=co_bls[co_bls['atype']=='CN'].set_index(['atype','fipscty','fipscty_str',\\\n", " 'indic','AUDIT_YEAR']).unstack(level='indic')\n", "\n", "#Rename columns (drop val level)\n", "co_labor_temp.columns=['emp','labforce','unemp','unemp_rate']\n", "\n", "#Reset index\n", "co_labor=co_labor_temp.reset_index()\n", "\n", "HTML(co_labor.head(10).to_html())" ], "language": "python", "metadata": {}, "outputs": [ { "html": [ "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "

	atype	fipscty	fipscty_str	AUDIT_YEAR	emp	labforce	unemp	unemp_rate
0	CN	1	001	1990	132892	140949	8057	5.7
1	CN	1	001	1991	134868	142668	7800	5.5
2	CN	1	001	1992	137948	146510	8562	5.8
3	CN	1	001	1993	143050	151092	8042	5.3
4	CN	1	001	1994	151859	158830	6971	4.4
5	CN	1	001	1995	159186	165754	6568	4.0
6	CN	1	001	1996	162391	169548	7157	4.2
7	CN	1	001	1997	167950	173611	5661	3.3
8	CN	1	001	1998	174580	180636	6056	3.4
9	CN	1	001	1999	178407	183583	5176	2.8

" ], "metadata": {}, "output_type": "pyout", "prompt_number": 30, "text": [ "" ] } ], "prompt_number": 30 }, { "cell_type": "code", "collapsed": false, "input": [ "state_unemp_dict={1987:7.5,\n", " 1988:6.4,\n", " 1989:5.6,\n", " 1990:5.1,\n", " 1991:5.4,\n", " 1992:6.0,\n", " 1993:5.3,\n", " 1994:4.3,\n", " 1995:4.0,\n", " 1996:4.2,\n", " 1997:3.4,\n", " 1998:3.5,\n", " 1999:3.0,\n", " 2000:2.7,\n", " 2001:3.8,\n", " 2002:5.7,\n", " 2003:6.1,\n", " 2004:5.6,\n", " 2005:5.1,\n", " 2006:4.3,\n", " 2007:3.8,\n", " 2008:4.8,\n", " 2009:8.1,\n", " 2010:9.0,\n", " 2011:8.6,\n", " 2012:8.0}\n", " " ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 31 }, { "cell_type": "markdown", "metadata": {}, "source": [ "We can now integrate the labor force information, integrate statewide unemployment information, and conclude the primary data management portion of the script." ] }, { "cell_type": "code", "collapsed": false, "input": [ "#Merge labor component in\n", "co_sfdal=pd.merge(co_sfda.reset_index(),co_labor,how='left',on=['fipscty','AUDIT_YEAR'])\n", "\n", "#Include state unemployment rate\n", "co_sfdal['st_unempr']=co_sfdal['AUDIT_YEAR'].map(state_unemp_dict)\n", "\n", "co_sfdal\n", "#HTML(co_sfdal.describe().T.to_html())" ], "language": "python", "metadata": {}, "outputs": [ { "html": [ "

\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "

	level_0	Unnamed: 0	ALAND	AWATER	CBSAFP	CLASSFP	COUNTYFP	COUNTYNS	CSAFP	FUNCSTAT	GEOID	INTPTLAT	INTPTLON	LSAD	METDIVFP	MTFCC	NAME	NAMELSAD	STATEFP	sort_id
0	0	0	2893573835	15324680	NaN	H1	53	198142	NaN	A	8053	37.811667	-107.383335	6	NaN	G4020	Hinsdale	Hinsdale County	8	0	...
1	1	1	2957922894	16953424	NaN	H1	29	198130	NaN	A	8029	38.861756	-107.864757	6	NaN	G4020	Delta	Delta County	8	1	...
2	2	2	388229141	941898	19740	H1	47	198139	216	A	8047	39.861082	-105.528947	6	NaN	G4020	Gilpin	Gilpin County	8	2	...
3	3	3	3177803460	8828904	NaN	H1	23	198127	NaN	A	8023	37.277547	-105.428940	6	NaN	G4020	Costilla	Costilla County	8	3	...
4	4	4	4179563675	18750018	NaN	H1	57	198144	NaN	A	8057	40.663432	-106.329248	6	NaN	G4020	Jackson	Jackson County	8	4	...
5	5	5	2624702134	4013482	NaN	H1	15	198123	NaN	A	8015	38.738244	-106.316955	6	NaN	G4020	Chaffee	Chaffee County	8	5	...
6	6	6	4793688729	442148	19740	H1	39	198136	216	A	8039	39.310582	-104.117729	6	NaN	G4020	Elbert	Elbert County	8	6	...
7	7	7	1442738671	4944573	17820	H1	119	198175	NaN	A	8119	38.871994	-105.182552	6	NaN	G4020	Teller	Teller County	8	7	...
8	8	8	4243419913	15352633	NaN	H1	99	198165	NaN	A	8099	37.958181	-102.392161	6	NaN	G4020	Prowers	Prowers County	8	8	...
9	9	9	1419419287	3530746	NaN	H1	115	198173	NaN	A	8115	40.871568	-102.355358	6	NaN	G4020	Sedgwick	Sedgwick County	8	9	...
10	10	10	396198354	4222992	19740	H6	31	198131	216	C	8031	39.761850	-104.880641	6	NaN	G4020	Denver	Denver County	8	10	...
11	11	11	2361929110	984917	NaN	H1	105	198168	NaN	A	8105	37.485763	-106.453214	6	NaN	G4020	Rio Grande	Rio Grande County	8	11	...
12	12	12	5256407750	27516850	NaN	H1	83	198157	NaN	A	8083	37.338025	-108.595786	6	NaN	G4020	Montezuma	Montezuma County	8	12	...
13	13	13	12284921637	19806932	NaN	H1	81	198156	NaN	A	8081	40.573984	-108.204521	6	NaN	G4020	Moffat	Moffat County	8	13	...
14	14	14	3334263504	9235291	NaN	H1	21	198126	NaN	A	8021	37.213406	-106.176447	6	NaN	G4020	Conejos	Conejos County	8	14	...
15	15	15	4382658416	19545450	20420	H1	67	198148	NaN	A	8067	37.287367	-107.839718	6	NaN	G4020	La Plata	La Plata County	8	15	...
16	16	16	2763645279	2716141	NaN	H1	33	198132	NaN	A	8033	37.747602	-108.530383	6	NaN	G4020	Dolores	Dolores County	8	16	...
17	17	17	5596501541	2251930	NaN	H1	63	198147	NaN	A	8063	39.305340	-102.603023	6	NaN	G4020	Kit Carson	Kit Carson County	8	17	...
18	18	18	10326885138	76675362	24540	H1	123	198177	216	A	8123	40.555794	-104.383649	6	NaN	G4020	Weld	Weld County	8	18	...
19	19	19	4578600049	47032937	NaN	H1	61	198146	NaN	A	8061	38.388466	-102.756210	6	NaN	G4020	Kiowa	Kiowa County	8	19	...
20	20	20	8207226426	4454545	NaN	H1	109	198170	NaN	A	8109	38.031651	-106.234666	6	NaN	G4020	Saguache	Saguache County	8	20	...
21	21	21	6521817532	15207414	NaN	H1	121	198176	NaN	A	8121	39.965790	-103.209744	6	NaN	G4020	Washington	Washington County	8	21	...
22	22	22	3496904695	13827548	NaN	H1	7	198119	NaN	A	8007	37.202395	-107.050863	6	NaN	G4020	Archuleta	Archuleta County	8	22	...
23	23	23	6617316468	6142193	NaN	H1	9	198120	NaN	A	8009	37.306678	-102.537765	6	NaN	G4020	Baca	Baca County	8	23	...
24	24	24	2176207910	6769100	19740	H1	35	198133	216	A	8035	39.326435	-104.926199	6	NaN	G4020	Douglas	Douglas County	8	24	...
25	25	25	6179976224	30284066	39380	H1	101	198166	NaN	A	8101	38.170658	-104.489893	6	NaN	G4020	Pueblo	Pueblo County	8	25	...
26	26	26	4362916056	18803966	20780	H1	37	198134	NaN	A	8037	39.630638	-106.692944	6	NaN	G4020	Eagle	Eagle County	8	26	...
27	27	27	1781724999	301808	NaN	H1	95	198163	NaN	A	8095	40.594712	-102.345105	6	NaN	G4020	Phillips	Phillips County	8	27	...
28	28	28	8342198615	4868111	NaN	H1	103	198167	NaN	A	8103	39.972606	-108.200685	6	NaN	G4020	Rio Blanco	Rio Blanco County	8	28	...
29	29	29	6117633825	15831696	NaN	H1	107	198169	NaN	A	8107	40.483155	-106.985274	6	NaN	G4020	Routt	Routt County	8	29	...
30	30	30	1575562624	28287714	43540	H1	117	198174	NaN	A	8117	39.621023	-106.137554	6	NaN	G4020	Summit	Summit County	8	30	...
31	31	31	3024201147	41944233	19740	H1	1	198116	216	A	8001	39.874325	-104.331872	6	NaN	G4020	Adams	Adams County	8	31	...
32	32	32	4762050035	16363552	44540	H1	75	198153	NaN	A	8075	40.728091	-103.090464	6	NaN	G4020	Logan	Logan County	8	32	...
33	33	33	5803398598	4923261	33940	H1	85	198158	NaN	A	8085	38.413427	-108.263042	6	NaN	G4020	Montrose	Montrose County	8	33	...
34	34	34	7634471698	21430626	NaN	H1	45	198138	NaN	A	8045	39.599352	-107.909780	6	NaN	G4020	Garfield	Garfield County	8	34	...
35	35	35	5508460909	7111548	17820	H1	41	198135	NaN	A	8041	38.827383	-104.527472	6	NaN	G4020	El Paso	El Paso County	8	35	...
36	36	36	2268230027	5222170	NaN	H1	79	198155	NaN	A	8079	37.651478	-106.932300	6	NaN	G4020	Mineral	Mineral County	8	36	...
37	37	37	5682134218	43530467	19740	H1	93	198162	216	A	8093	39.118914	-105.717648	6	NaN	G4020	Park	Park County	8	37	...
38	38	38	4120753930	5792107	NaN	H1	55	198143	NaN	A	8055	37.687815	-104.959928	6	NaN	G4020	Huerfano	Huerfano County	8	38	...
39	39	39	3918307819	73093700	NaN	H1	11	198121	NaN	A	8011	37.931891	-103.077584	6	NaN	G4020	Bent	Bent County	8	39	...
40	40	40	1023554447	3280081	19740	H1	19	198125	216	A	8019	39.689402	-105.670791	6	NaN	G4020	Clear Creek	Clear Creek County	8	40	...
41	41	41	976206766	18110615	20780	H1	65	198149	NaN	A	8065	39.205341	-106.350079	6	NaN	G4020	Lake	Lake County	8	41	...
42	42	42	4605714106	8166145	NaN	H1	17	198124	NaN	A	8017	38.835386	-102.604585	6	NaN	G4020	Cheyenne	Cheyenne County	8	42	...
43	43	43	12361253683	6929976	NaN	H1	71	198151	NaN	A	8071	37.318831	-104.044110	6	NaN	G4020	Las Animas	Las Animas County	8	43	...
44	44	44	1913033457	3364740	NaN	H1	27	198129	NaN	A	8027	38.101994	-105.373515	6	NaN	G4020	Custer	Custer County	8	44	...
45	45	45	1871620246	1847609	NaN	H1	3	198117	NaN	A	8003	37.568442	-105.788041	6	NaN	G4020	Alamosa	Alamosa County	8	45	...
46	46	46	6123751576	11134682	NaN	H1	125	198178	NaN	A	8125	40.000631	-102.422649	6	NaN	G4020	Yuma	Yuma County	8	46	...
47	47	47	8622131029	31435336	24300	H1	77	198154	NaN	A	8077	39.019492	-108.461837	6	NaN	G4020	Mesa	Mesa County	8	47	...
48	48	48	4781807818	60250181	NaN	H1	49	198140	NaN	A	8049	40.123289	-106.095876	6	NaN	G4020	Grand	Grand County	8	48	...
49	49	49	1402725842	1599543	NaN	H1	91	198161	NaN	A	8091	38.150600	-107.767133	6	NaN	G4020	Ouray	Ouray County	8	49	...
50	50	50	8389491951	53067063	NaN	H1	51	198141	NaN	A	8051	38.669679	-107.078108	6	NaN	G4020	Gunnison	Gunnison County	8	50	...
51	51	51	3332207690	5213860	NaN	H1	113	198172	NaN	A	8113	38.009373	-108.427326	6	NaN	G4020	San Miguel	San Miguel County	8	51	...
52	52	52	3316112929	34663780	22820	H1	87	198159	NaN	A	8087	40.262540	-103.807497	6	NaN	G4020	Morgan	Morgan County	8	52	...
53	53	53	6723779313	98113981	22660	H1	69	198150	NaN	A	8069	40.663091	-105.482131	6	NaN	G4020	Larimer	Larimer County	8	53	...
54	54	54	3268485084	20052353	NaN	H1	89	198160	NaN	A	8089	37.884170	-103.721260	6	NaN	G4020	Otero	Otero County	8	54	...
55	55	55	2514071663	6469378	NaN	H1	97	198164	NaN	A	8097	39.217533	-106.915943	6	NaN	G4020	Pitkin	Pitkin County	8	55	...
56	56	56	6676208280	22780444	NaN	H1	73	198152	NaN	A	8073	38.993740	-103.507554	6	NaN	G4020	Lincoln	Lincoln County	8	56	...
57	57	57	2039409780	33430387	NaN	H1	25	198128	NaN	A	8025	38.306180	-103.772736	6	NaN	G4020	Crowley	Crowley County	8	57	...
58	58	58	1003640124	2035929	NaN	H1	111	198171	NaN	A	8111	37.781075	-107.670257	6	NaN	G4020	San Juan	San Juan County	8	58	...
59	59	59	3970604102	2235391	15860	H1	43	198137	NaN	A	8043	38.455658	-105.421438	6	NaN	G4020	Fremont	Fremont County	8	59	...
	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...

\n", "

1472 rows \u00d7 164 columns

\n", "

" ], "metadata": {}, "output_type": "pyout", "prompt_number": 32, "text": [ " level_0 Unnamed: 0 ALAND AWATER CBSAFP CLASSFP COUNTYFP COUNTYNS CSAFP FUNCSTAT GEOID INTPTLAT INTPTLON LSAD \\\n", "0 0 0 2893573835 15324680 NaN H1 53 198142 NaN A 8053 37.811667 -107.383335 6 \n", "1 1 1 2957922894 16953424 NaN H1 29 198130 NaN A 8029 38.861756 -107.864757 6 \n", "2 2 2 388229141 941898 19740 H1 47 198139 216 A 8047 39.861082 -105.528947 6 \n", "3 3 3 3177803460 8828904 NaN H1 23 198127 NaN A 8023 37.277547 -105.428940 6 \n", "4 4 4 4179563675 18750018 NaN H1 57 198144 NaN A 8057 40.663432 -106.329248 6 \n", "5 5 5 2624702134 4013482 NaN H1 15 198123 NaN A 8015 38.738244 -106.316955 6 \n", "6 6 6 4793688729 442148 19740 H1 39 198136 216 A 8039 39.310582 -104.117729 6 \n", "7 7 7 1442738671 4944573 17820 H1 119 198175 NaN A 8119 38.871994 -105.182552 6 \n", "8 8 8 4243419913 15352633 NaN H1 99 198165 NaN A 8099 37.958181 -102.392161 6 \n", "9 9 9 1419419287 3530746 NaN H1 115 198173 NaN A 8115 40.871568 -102.355358 6 \n", "10 10 10 396198354 4222992 19740 H6 31 198131 216 C 8031 39.761850 -104.880641 6 \n", "11 11 11 2361929110 984917 NaN H1 105 198168 NaN A 8105 37.485763 -106.453214 6 \n", "12 12 12 5256407750 27516850 NaN H1 83 198157 NaN A 8083 37.338025 -108.595786 6 \n", "13 13 13 12284921637 19806932 NaN H1 81 198156 NaN A 8081 40.573984 -108.204521 6 \n", "14 14 14 3334263504 9235291 NaN H1 21 198126 NaN A 8021 37.213406 -106.176447 6 \n", "15 15 15 4382658416 19545450 20420 H1 67 198148 NaN A 8067 37.287367 -107.839718 6 \n", "16 16 16 2763645279 2716141 NaN H1 33 198132 NaN A 8033 37.747602 -108.530383 6 \n", "17 17 17 5596501541 2251930 NaN H1 63 198147 NaN A 8063 39.305340 -102.603023 6 \n", "18 18 18 10326885138 76675362 24540 H1 123 198177 216 A 8123 40.555794 -104.383649 6 \n", "19 19 19 4578600049 47032937 NaN H1 61 198146 NaN A 8061 38.388466 -102.756210 6 \n", "20 20 20 8207226426 4454545 NaN H1 109 198170 NaN A 8109 38.031651 -106.234666 6 \n", "21 21 21 6521817532 15207414 NaN H1 121 198176 NaN A 8121 39.965790 -103.209744 6 \n", "22 22 22 3496904695 13827548 NaN H1 7 198119 NaN A 8007 37.202395 -107.050863 6 \n", "23 23 23 6617316468 6142193 NaN H1 9 198120 NaN A 8009 37.306678 -102.537765 6 \n", "24 24 24 2176207910 6769100 19740 H1 35 198133 216 A 8035 39.326435 -104.926199 6 \n", "25 25 25 6179976224 30284066 39380 H1 101 198166 NaN A 8101 38.170658 -104.489893 6 \n", "26 26 26 4362916056 18803966 20780 H1 37 198134 NaN A 8037 39.630638 -106.692944 6 \n", "27 27 27 1781724999 301808 NaN H1 95 198163 NaN A 8095 40.594712 -102.345105 6 \n", "28 28 28 8342198615 4868111 NaN H1 103 198167 NaN A 8103 39.972606 -108.200685 6 \n", "29 29 29 6117633825 15831696 NaN H1 107 198169 NaN A 8107 40.483155 -106.985274 6 \n", "30 30 30 1575562624 28287714 43540 H1 117 198174 NaN A 8117 39.621023 -106.137554 6 \n", "31 31 31 3024201147 41944233 19740 H1 1 198116 216 A 8001 39.874325 -104.331872 6 \n", "32 32 32 4762050035 16363552 44540 H1 75 198153 NaN A 8075 40.728091 -103.090464 6 \n", "33 33 33 5803398598 4923261 33940 H1 85 198158 NaN A 8085 38.413427 -108.263042 6 \n", "34 34 34 7634471698 21430626 NaN H1 45 198138 NaN A 8045 39.599352 -107.909780 6 \n", "35 35 35 5508460909 7111548 17820 H1 41 198135 NaN A 8041 38.827383 -104.527472 6 \n", "36 36 36 2268230027 5222170 NaN H1 79 198155 NaN A 8079 37.651478 -106.932300 6 \n", "37 37 37 5682134218 43530467 19740 H1 93 198162 216 A 8093 39.118914 -105.717648 6 \n", "38 38 38 4120753930 5792107 NaN H1 55 198143 NaN A 8055 37.687815 -104.959928 6 \n", "39 39 39 3918307819 73093700 NaN H1 11 198121 NaN A 8011 37.931891 -103.077584 6 \n", "40 40 40 1023554447 3280081 19740 H1 19 198125 216 A 8019 39.689402 -105.670791 6 \n", "41 41 41 976206766 18110615 20780 H1 65 198149 NaN A 8065 39.205341 -106.350079 6 \n", "42 42 42 4605714106 8166145 NaN H1 17 198124 NaN A 8017 38.835386 -102.604585 6 \n", "43 43 43 12361253683 6929976 NaN H1 71 198151 NaN A 8071 37.318831 -104.044110 6 \n", "44 44 44 1913033457 3364740 NaN H1 27 198129 NaN A 8027 38.101994 -105.373515 6 \n", "45 45 45 1871620246 1847609 NaN H1 3 198117 NaN A 8003 37.568442 -105.788041 6 \n", "46 46 46 6123751576 11134682 NaN H1 125 198178 NaN A 8125 40.000631 -102.422649 6 \n", "47 47 47 8622131029 31435336 24300 H1 77 198154 NaN A 8077 39.019492 -108.461837 6 \n", "48 48 48 4781807818 60250181 NaN H1 49 198140 NaN A 8049 40.123289 -106.095876 6 \n", "49 49 49 1402725842 1599543 NaN H1 91 198161 NaN A 8091 38.150600 -107.767133 6 \n", "50 50 50 8389491951 53067063 NaN H1 51 198141 NaN A 8051 38.669679 -107.078108 6 \n", "51 51 51 3332207690 5213860 NaN H1 113 198172 NaN A 8113 38.009373 -108.427326 6 \n", "52 52 52 3316112929 34663780 22820 H1 87 198159 NaN A 8087 40.262540 -103.807497 6 \n", "53 53 53 6723779313 98113981 22660 H1 69 198150 NaN A 8069 40.663091 -105.482131 6 \n", "54 54 54 3268485084 20052353 NaN H1 89 198160 NaN A 8089 37.884170 -103.721260 6 \n", "55 55 55 2514071663 6469378 NaN H1 97 198164 NaN A 8097 39.217533 -106.915943 6 \n", "56 56 56 6676208280 22780444 NaN H1 73 198152 NaN A 8073 38.993740 -103.507554 6 \n", "57 57 57 2039409780 33430387 NaN H1 25 198128 NaN A 8025 38.306180 -103.772736 6 \n", "58 58 58 1003640124 2035929 NaN H1 111 198171 NaN A 8111 37.781075 -107.670257 6 \n", "59 59 59 3970604102 2235391 15860 H1 43 198137 NaN A 8043 38.455658 -105.421438 6 \n", " ... ... ... ... ... ... ... ... ... ... ... ... ... ... \n", "\n", " METDIVFP MTFCC NAME NAMELSAD STATEFP sort_id \n", "0 NaN G4020 Hinsdale Hinsdale County 8 0 ... \n", "1 NaN G4020 Delta Delta County 8 1 ... \n", "2 NaN G4020 Gilpin Gilpin County 8 2 ... \n", "3 NaN G4020 Costilla Costilla County 8 3 ... \n", "4 NaN G4020 Jackson Jackson County 8 4 ... \n", "5 NaN G4020 Chaffee Chaffee County 8 5 ... \n", "6 NaN G4020 Elbert Elbert County 8 6 ... \n", "7 NaN G4020 Teller Teller County 8 7 ... \n", "8 NaN G4020 Prowers Prowers County 8 8 ... \n", "9 NaN G4020 Sedgwick Sedgwick County 8 9 ... \n", "10 NaN G4020 Denver Denver County 8 10 ... \n", "11 NaN G4020 Rio Grande Rio Grande County 8 11 ... \n", "12 NaN G4020 Montezuma Montezuma County 8 12 ... \n", "13 NaN G4020 Moffat Moffat County 8 13 ... \n", "14 NaN G4020 Conejos Conejos County 8 14 ... \n", "15 NaN G4020 La Plata La Plata County 8 15 ... \n", "16 NaN G4020 Dolores Dolores County 8 16 ... \n", "17 NaN G4020 Kit Carson Kit Carson County 8 17 ... \n", "18 NaN G4020 Weld Weld County 8 18 ... \n", "19 NaN G4020 Kiowa Kiowa County 8 19 ... \n", "20 NaN G4020 Saguache Saguache County 8 20 ... \n", "21 NaN G4020 Washington Washington County 8 21 ... \n", "22 NaN G4020 Archuleta Archuleta County 8 22 ... \n", "23 NaN G4020 Baca Baca County 8 23 ... \n", "24 NaN G4020 Douglas Douglas County 8 24 ... \n", "25 NaN G4020 Pueblo Pueblo County 8 25 ... \n", "26 NaN G4020 Eagle Eagle County 8 26 ... \n", "27 NaN G4020 Phillips Phillips County 8 27 ... \n", "28 NaN G4020 Rio Blanco Rio Blanco County 8 28 ... \n", "29 NaN G4020 Routt Routt County 8 29 ... \n", "30 NaN G4020 Summit Summit County 8 30 ... \n", "31 NaN G4020 Adams Adams County 8 31 ... \n", "32 NaN G4020 Logan Logan County 8 32 ... \n", "33 NaN G4020 Montrose Montrose County 8 33 ... \n", "34 NaN G4020 Garfield Garfield County 8 34 ... \n", "35 NaN G4020 El Paso El Paso County 8 35 ... \n", "36 NaN G4020 Mineral Mineral County 8 36 ... \n", "37 NaN G4020 Park Park County 8 37 ... \n", "38 NaN G4020 Huerfano Huerfano County 8 38 ... \n", "39 NaN G4020 Bent Bent County 8 39 ... \n", "40 NaN G4020 Clear Creek Clear Creek County 8 40 ... \n", "41 NaN G4020 Lake Lake County 8 41 ... \n", "42 NaN G4020 Cheyenne Cheyenne County 8 42 ... \n", "43 NaN G4020 Las Animas Las Animas County 8 43 ... \n", "44 NaN G4020 Custer Custer County 8 44 ... \n", "45 NaN G4020 Alamosa Alamosa County 8 45 ... \n", "46 NaN G4020 Yuma Yuma County 8 46 ... \n", "47 NaN G4020 Mesa Mesa County 8 47 ... \n", "48 NaN G4020 Grand Grand County 8 48 ... \n", "49 NaN G4020 Ouray Ouray County 8 49 ... \n", "50 NaN G4020 Gunnison Gunnison County 8 50 ... \n", "51 NaN G4020 San Miguel San Miguel County 8 51 ... \n", "52 NaN G4020 Morgan Morgan County 8 52 ... \n", "53 NaN G4020 Larimer Larimer County 8 53 ... \n", "54 NaN G4020 Otero Otero County 8 54 ... \n", "55 NaN G4020 Pitkin Pitkin County 8 55 ... \n", "56 NaN G4020 Lincoln Lincoln County 8 56 ... \n", "57 NaN G4020 Crowley Crowley County 8 57 ... \n", "58 NaN G4020 San Juan San Juan County 8 58 ... \n", "59 NaN G4020 Fremont Fremont County 8 59 ... \n", " ... ... ... ... ... ... \n", "\n", "[1472 rows x 164 columns]" ] } ], "prompt_number": 32 }, { "cell_type": "heading", "level": 2, "metadata": {}, "source": [ "Model Specification" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The model used in the initial stab (*CO_TEL_convergence*) at explaining variance in revenue yield was very limited, both in terms of the regressors included and the explanatory power of the model. Rehashing this analysis provides an opportunity to think more deeply about the relationships we may exploit. \n", "\n", "In general, there are three phenomenon we wish to explain as a function of TEL intensity: \n", "\n", "1. **Economic Capacity** - *Regardless of the explicit fiscal choices made, what is the economic carrying capacity within a given jurisdiction?*\n", "\n", " + Ultimately, the dynamic impact of TELs on economic growth is the top line goal of this inquiry. This is a somewhat brute force method of exploring that impact, but it is likely to blunt an instrument to sufficiently capture the causal mechanism.\n", " + Additional Regressors:\n", " + Previous year's capacity\n", " + Weighted average of local neighborhood's capacities\n", " + Gross state product\n", " + Business cycle \n", " + State unemployment rate\n", " + Demand for housing in jurisdiction\n", " + Housing permits/Housing units\n", " + Vacancy rate\n", " + Labor concentration \n", " + Labor force/Population (for now, incomplete county coverage prohibits inclusion)\n", "\n", "2. **Revenue Yield** - *How much revenue is generated per capita in a given jurisdiction?*\n", "\n", " + There are actually two ways to think about this. The most direct method would be as an autonomous model, that may then be compared to the economic capacity model. The second method seeks to integrate both economic capacity and revenue yield via a hierarchical model. Given the \"carrying capacity\" of a given jurisdiction, the yield per capita may be represented as tax effort.\n", " + Additional Regressors:\n", " + Previous year's revenue\n", " + Weighted average of local neighborhood's yields\n", " + Gross state product\n", " + Demand for services\n", " + Lagged natural population growth\n", " + Business cycle \n", " + State unemployment rate\n", " + Demand for housing in jurisdiction\n", " + Housing permits/Housing units\n", " + Vacancy rate\n", " + Economic capacity\n", " + Annual payroll per capita\n", " + Labor concentration\n", " + Labor force/Population (for now, incomplete county coverage prohibits inclusion)\n", " \n", "3. **Spatial Clustering** - *How much does a given jurisdiction look like its local neighborhood?*\n", "\n", " + From an operational standpoint, we want to know 1) if TELs cause a jurisdiction to look more or less like it's local neighborhood, and 2) if TELs are more likely to cause clustering among low capacity jurisdictions or high capacity jurisdictions. Local Moran's I and Getis & Ord's G* estimates provide the means to evaluate these statements, respectively. This would, of course, need to be run for both contiguity and distance based weight matrices.\n", " + Additional Regressors:\n", " + Previous year's clustering value \n", " + Gross state product\n", " + Demand for services\n", " + Lagged natural population growth\n", " + Business cycle \n", " + State unemployment rate\n", " + Demand for housing in jurisdiction and the local neighborhood\n", " + Housing permits/Housing units\n", " + Vacancy rate\n", " + Economic capacity\n", " + Annual payroll per capita\n", " + Labor concentration\n", " + Labor force/Population (for now, incomplete county coverage prohibits inclusion)\n", " \n", "Furthermore, we must explore whether or not these behaviors change with different economic starting conditions. Consequently, after the initial runs on the full set, we will run the same analyses on subsets of the data split by per capita annual payroll quintiles in year 0." ] }, { "cell_type": "heading", "level": 4, "metadata": {}, "source": [ "TEL Intensity" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We also need to develop a method of evaluating TEL intensity on a given county. To borrow from the *tabor_score* Notebook, measuring the impact of TELs requires four components:\n", "\n", "\n", "1. Comparison between the growth in the total property tax base (residential and commercial) with the statewide property tax growth limitation of 5.5% annually. We can measure the rate of growth of the assessment base.\n", " + Total Property Tax Base Growth = $\\frac{resid_t + non\\_resid_t}{resid_{t-1} + non\\_resid_{t-1}} - 1$\n", "2. Comparison between the growth in total property base and the percentage growth value attributable to inflation plus new construction. (Note that the growth factor for school districts is actually enrollment figures, so this measure would be imperfect pending greater jurisdictional resolution.) This is quite similar to the first component, except that the limit is variable instead of fixed at 5.5%. In practice, the composite limit (Statewide Property Tax Growth Limitation and TABOR) is governed by this relationship below. This limit is modified only when De-Brucing occurs. De-Brucing may lead the limit to reduce to one of the two arguments in the minimum function, or to disappear entirely. Either way, the property tax growth is calculated in the same way, and then compared with whichever version of the limit exists for a given county.\n", "\n", " $$PRlimit = min[.055,inflation + construction]$$\n", " \n", "\n", "3. Comparison the market value growth in residential and non-residential property tax base. This is needed for both the state and county levels. The former is needed only to understand the shift in assessment rates (which are set statewide), and thus could be circumvented by just getting the assessment rates by year. The latter is critical so that we may be able to see the impact of drops in the residential assessment rate by county. One way to get at this concept would be to compare the ratio of residential to non-residential assessment bases across years. An increase in the ratio would indicate greater growth in residential assessment value relative to non-residential. \n", "\n", "$$ASMT\\_RATIO = \\frac{Res\\_ASMT}{NonRes\\_ASMT} = \\frac{Res\\_VALUE * Res\\_RATE}{NonRes\\_VALUE * NonRes\\_RATE}$$\n", "\n", "Maybe I just want the ratio because it tells me about residential concentration as a characteristic of the county. A residential concentration increases the difficulty of matching revenue growth to the increase in the property-based activity that might benefit from public provision of infrastructure and other services.\n", "\n", "Since equivalent increases in market value growth yield slower growth for residential assessment value\n", "\n", "4. State aid allocations by county. These should be available from budgetary information.\n", "\n", "To validate these as measures, it would be useful to explore the dynamics of these components over time." ] }, { "cell_type": "heading", "level": 4, "metadata": {}, "source": [ "Assessment Base Growth" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The first step is calculating the base growth for residential, non-residential, and total property assessment levels. Note that we want to avoid the percentage change calculation at the boundary between counties. Enter the old split-apply-combine. To keep within county integrity, we will do the following:\n", "\n", "1. Split the data by county\n", "2. Calculate the percentage change\n", "3. Replace any *inf* with *NaN* (*inf* occurs when a zero value transitions to a non-zero value)\n", "4. Concatenate the DF pieces back together into one large DF\n", "\n", "We will need housing unit growth later in the script as well, so we might as well calculate that now." ] }, { "cell_type": "code", "collapsed": false, "input": [ "co_sfdal.columns" ], "language": "python", "metadata": {}, "outputs": [ { "metadata": {}, "output_type": "pyout", "prompt_number": 33, "text": [ "Index([u'level_0', u'Unnamed: 0', u'ALAND', u'AWATER', u'CBSAFP', u'CLASSFP', u'COUNTYFP', u'COUNTYNS', u'CSAFP', u'FUNCSTAT', u'GEOID', u'INTPTLAT', u'INTPTLON', u'LSAD', u'METDIVFP', u'MTFCC', u'NAME', u'NAMELSAD', u'STATEFP', u'sort_id', u'index', u'FNAME', u'AUDIT_YEAR', u'REV_TOTAL', u'REV_INTGOVT', u'EXP_TOTAL', u'TOTAL_DEBT', u'CTY_POP', u'intensity', u'pcrev', u'pcexp', u'pcintgov', u'pcdebt', u'pc_ap', u'w_rook_i', u'w_queen_i', u'w_db_b_i', u'w_db_c_i', u'w_kern_i', u'w_rook_i_ap', u'w_queen_i_ap', u'w_db_b_i_ap', u'w_db_c_i_ap', u'w_kern_i_ap', u'w_rook_g', u'w_queen_g', u'w_db_b_g', u'w_db_c_g', u'w_kern_g', u'w_rook_g_ap', u'w_queen_g_ap', u'w_db_b_g_ap', u'w_db_c_g_ap', u'w_kern_g_ap', u'fipsc', u'fipscty', u'year_x', u'ap', u'pcap', u'pcap_mean', u'pcap_rel', u'gsp', u'pcap_rel2', u'LG_ID', u'REV_CTF', u'EXP_DEBT_SERVICE_GEN', u'EXP_TRANSFER_OUT', u'ST_SALES_TAX_PAID', u'EXP_CAPITAL_OUTLAY', u'EXP_JUDICIAL', u'db', u'EXP_HEALTH', u'EXP_FIRE', u'EXP_OTHER_PUBLIC_WORKS', u'REV_FRANCHISE_TAX', u'POPULATION', u'EXP_GEN_GOVT', u'LIABILITIES', u'EXP_TOTAL_PUBLIC_SAFETY', u'REV_LODGING_TAX', u'ALL_OTHER_INTGOVT', u'REVENUE_DEBT_GEN', u'EXP_MISC', u'REV_OTHER_TAX', u'REV_UNCLASS_TAX', u'db_temp', u'GO_DEBT_GEN', u'OTHER_DEBT_GEN', u'sorting_id', u'EXP_SOCIAL_SERVICE', u'EXP_PRINCIPAL_GEN', u'REV_MOTOR_VEH_FEE', u'REV_TRANSFER_IN', u'EXP_RECREATION', u'EXP_POLICE', u'EXP_INTEREST_GEN', u'NAME_x', u'NAME_y', u'EXP_OTHER_PUBLIC_SAFETY', u'REV_CHARGES', ...], dtype='object')" ] } ], "prompt_number": 33 }, { "cell_type": "code", "collapsed": false, "input": [ "#Subset to after 1992 (note that all change variables, including assessment, will be NaN in 1992)\n", "co_92=co_sfdal[co_sfdal['AUDIT_YEAR']>1991].set_index(['FNAME','AUDIT_YEAR']).sortlevel(0)\n", "\n", "#Create containers for DF components\n", "co_DF_list=[]\n", "\n", "#For every county...\n", "for cty in set(co_92.reset_index()['FNAME']):\n", " #Create a county subset\n", " co_cty=co_92.ix[cty]\n", " #Create percentage change variables\n", " co_cty['d_resid']=co_cty['resid'].pct_change()\n", " co_cty['d_non_resid']=co_cty['non_resid'].pct_change()\n", " co_cty['d_total']=co_cty['total'].pct_change()\n", " co_cty['hu_growth']=co_cty['hu_num'].pct_change()\n", " co_cty['pop_growth']=co_cty['CTY_POP'].pct_change()\n", " #Replace any resultant 'inf' with 'NaN'\n", " co_cty2=co_cty.replace(inf,NaN)\n", " #Reset county name\n", " co_cty2['FNAME']=cty\n", " #Throw it back in the container\n", " co_DF_list.append(co_cty2)\n", " \n", "#Concatenate pieces back together\n", "co92_temp=pd.concat(co_DF_list).reset_index()\n", "\n", "#Set index again\n", "co93=co92_temp[co92_temp['AUDIT_YEAR']>1992].set_index(['FNAME','AUDIT_YEAR']).sortlevel(0)\n", " \n", "print co93[['resid','d_resid','non_resid','d_non_resid','total','d_total']].head(10)\n", "co93[['resid','d_resid','non_resid','d_non_resid','total','d_total']].describe()" ], "language": "python", "metadata": {}, "outputs": [ { "output_type": "stream", "stream": "stdout", "text": [ " resid d_resid non_resid d_non_resid total d_total\n", "FNAME AUDIT_YEAR \n", "Adams 1993 705269720 NaN 927339630 NaN 1632609350 NaN\n", " 1994 742865190 0.053307 941530050 0.015302 1684395240 0.031720\n", " 1995 813385250 0.094930 1001215680 0.063392 1814600930 0.077301\n", " 1996 853089120 0.048813 1045276840 0.044008 1898365960 0.046162\n", " 1997 955203880 0.119700 1138878080 0.089547 2094081960 0.103097\n", " 1998 1003073550 0.050115 1201323630 0.054831 2204397180 0.052680\n", " 1999 1181599650 0.177979 1371347110 0.141530 2552946760 0.158116\n", " 2000 1257611690 0.064330 1448916800 0.056565 2706528490 0.060159\n", " 2001 1568125310 0.246907 1732989150 0.196058 3301114460 0.219686\n", " 2002 1581702560 0.008658 1761408110 0.016399 3343110670 0.012722\n", "\n", "[10 rows x 6 columns]\n" ] }, { "html": [ "

	resid	d_resid	non_resid	d_non_resid	total	d_total
count	1.088000e+03	1016.000000	1.088000e+03	1016.000000	1.088000e+03	1016.000000
mean	4.088715e+08	0.072506	4.739104e+08	0.076939	8.827819e+08	0.076841
std	8.365265e+08	0.088602	8.685433e+08	0.125075	1.659197e+09	0.106413
min	0.000000e+00	-0.194075	0.000000e+00	-0.236877	0.000000e+00	-0.220550
25%	1.317100e+07	0.024101	4.804648e+07	0.004394	6.657231e+07	0.015231
50%	4.965357e+07	0.050673	1.144799e+08	0.045165	1.815781e+08	0.046211
75%	2.813840e+08	0.106665	4.187223e+08	0.119831	7.443404e+08	0.118784
max	4.546922e+09	0.533770	7.476110e+09	1.008305	1.202303e+10	0.842019

\n", "

8 rows \u00d7 6 columns

\n", "

" ], "metadata": {}, "output_type": "pyout", "prompt_number": 34, "text": [ " resid d_resid non_resid d_non_resid total d_total\n", "count 1.088000e+03 1016.000000 1.088000e+03 1016.000000 1.088000e+03 1016.000000\n", "mean 4.088715e+08 0.072506 4.739104e+08 0.076939 8.827819e+08 0.076841\n", "std 8.365265e+08 0.088602 8.685433e+08 0.125075 1.659197e+09 0.106413\n", "min 0.000000e+00 -0.194075 0.000000e+00 -0.236877 0.000000e+00 -0.220550\n", "25% 1.317100e+07 0.024101 4.804648e+07 0.004394 6.657231e+07 0.015231\n", "50% 4.965357e+07 0.050673 1.144799e+08 0.045165 1.815781e+08 0.046211\n", "75% 2.813840e+08 0.106665 4.187223e+08 0.119831 7.443404e+08 0.118784\n", "max 4.546922e+09 0.533770 7.476110e+09 1.008305 1.202303e+10 0.842019\n", "\n", "[8 rows x 6 columns]" ] } ], "prompt_number": 34 }, { "cell_type": "markdown", "metadata": {}, "source": [ "To get a sense of how much assessment values move around, let's examine Adams County." ] }, { "cell_type": "code", "collapsed": false, "input": [ "plt.rcParams['figure.figsize']=15,5\n", "co93.ix['Adams'][['d_total','d_resid','d_non_resid']].plot(kind='bar',title='Assessment Base - Annual % Change')\n" ], "language": "python", "metadata": {}, "outputs": [ { "metadata": {}, "output_type": "pyout", "prompt_number": 35, "text": [ "" ] }, { "metadata": {}, "output_type": "display_data", "png": "iVBORw0KGgoAAAANSUhEUgAAA3QAAAFgCAYAAADzSDsfAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n", "text": [ "" ] } ], "prompt_number": 35 }, { "cell_type": "markdown", "metadata": {}, "source": [ "It is immediately apparent that the 5.5% statewide limit is a binding constraint with some frequency in Adams County. A boxplot of each of these indicators over time would provide some sense of how the overall distribution looks for all counties." ] }, { "cell_type": "code", "collapsed": false, "input": [ "plt.rcParams['figure.figsize']=15,10\n", "for i in range(1993,2010):\n", " y1=co93.xs(i,level='AUDIT_YEAR')['d_total'].dropna()\n", " y2=co93.xs(i,level='AUDIT_YEAR')['d_resid'].dropna()\n", " y3=co93.xs(i,level='AUDIT_YEAR')['d_non_resid'].dropna()\n", " x1=np.random.normal(i,.04,size=len(y1))\n", " x2=np.random.normal(i,.04,size=len(y2))\n", " x3=np.random.normal(i,.04,size=len(y3))\n", " plot(x1,y1,'r.',alpha=.2)\n", " plot(x2,y2,'b.',alpha=.2)\n", " plot(x3,y3,'g.',alpha=.2)" ], "language": "python", "metadata": {}, "outputs": [ { "metadata": {}, "output_type": "display_data", "png": "iVBORw0KGgoAAAANSUhEUgAAA3oAAAJPCAYAAAAqk6bSAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nhV4SYCl58LsAAAAASUVORK5CYII=\n", "text": [ "" ] } ], "prompt_number": 36 }, { "cell_type": "code", "collapsed": false, "input": [ "plt.rcParams['figure.figsize']=15,5\n", "plt.figure()\n", "co93.reset_index().boxplot(column='d_total',by='AUDIT_YEAR')\n", "plt.figure()\n", "co93.reset_index().boxplot(column='d_resid',by='AUDIT_YEAR')\n", "plt.figure()\n", "co93.reset_index().boxplot(column='d_non_resid',by='AUDIT_YEAR')" ], "language": "python", "metadata": {}, "outputs": [ { "metadata": {}, "output_type": "pyout", "prompt_number": 37, "text": [ "" ] }, { "metadata": {}, "output_type": "display_data", "text": [ "" ] }, { "metadata": {}, "output_type": "display_data", "png": "iVBORw0KGgoAAAANSUhEUgAAA4kAAAFXCAYAAAAPhQsjAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzt3Xt8FNX9//H3IlhR0CgoCEHD/WIgCSCKVl1LEWkLtaAW\nvCX1brUCfbQiVr+QfrVKy7cNlkf7o9ZKqjVqhYLXeMEMRWsANUFRtBUJBBBabnLRcgnz+yNmSUiy\nZndnd86cfT0fj9VMMrv7eXN2dvbsnDMTcl3XFQAAAAAAklr5XQAAAAAAwBx0EgEAAAAAEXQSAQAA\nAAARdBIBAAAAABF0EgEAAAAAEXQSAQAAAAARdBIBAAAAABF0EgEgzR111FHKy8tTbm6uhgwZojff\nfNPTx3ccR2PGjIm6zpIlSzx/3lTIysrS9u3bG/2+Xbt2CT3uwYMHdfLJJ2vatGlRn6/+v+28efN0\n8skna/DgwerTp48uvvjiBv+mBQUFmj9/vsaNG6e8vDz17t1bGRkZysvLU15ensrLyxvVcffdd+vO\nO++MLK9bt049e/bUZ599pnA4rH79+kXuf/nllze4b25uriZOnNjgdwUFBerRo4fy8vI0ePBgLV26\nNP5/JABA0rT2uwAAgL+OPfZYVVRUSJJefvllTZs2TY7jpLSGsrIytW/fXsOHD4/r/q7rSpJCoZCX\nZX2l5p4v0TpeeeUVDRkyRPPnz9f999/foscNhUKaOHGiHnzwQUm1Hchx48aprKxM/fr1UygUUigU\n0oIFCyTVdsxnzZqlZ599ttnHvPvuu5Wbm6uCggL169dPkyZN0r333qsTTjhBoVBIjz/+uAYPHtzo\nfqtXr9YxxxyjZcuW6fPPP9exxx4bqXHWrFkaN26cHMfRD3/4Q7333ntx/RsBAJKHI4kAgIjPPvtM\nJ510kqTajtdPf/pTDRw4UIMGDdJTTz0lSZo8ebL+93//V5L00ksv6YILLpDruiooKNDNN9+sM888\nU3379tXzzz/f6PG3b9+uSy65RDk5ORo+fLjee+89VVVVae7cufrNb36jvLw8vf766w3u85///Ecj\nR45Udna2brjhhsjRtKqqKvXt21f5+fkaOHCgqqurm6z3yCOZt912m4qLiyXVHpmbOnWqBg0apLPO\nOktr1qyJPOell16qYcOGadiwYfrHP/4hSdq2bZsuuuiiSC11ndOm/PjHP1Z2dra++c1vauvWrVqz\nZo2GDBkS+fu//vWvBsv1PfHEE7rlllvUo0ePFh9hdV23QT3hcFg33nij/vCHPzRYp6mfm3PMMcfo\nN7/5jW699Va98MIL2rt3b4Ojg809RklJiSZOnKiLLrpIixYtalSnJJ199tmRf28AgFnoJAJAmvvi\niy+Ul5en/v3764YbbtA999wjSVqwYIFWrlypd999V6+++qp++tOfasuWLbr//vv15JNPqqysTJMm\nTdK8efMiR7jWr1+vFStW6Pnnn9fNN9+sffv2NXiu6dOna8iQIVq5cqV+8Ytf6JprrlFWVpZuvvlm\n/fjHP1ZFRYW+/vWvN7hPYWGhvvnNb2rVqlW69NJLtX79+sjfPv74Y916661atWqVVqxY0ajezZs3\nN8pbd0St7ueMjAy9++67uu222zR58mRJ0qRJkzRlyhQtX75cTz/9tK6//vpILeeff75WrVql733v\new1qqW/v3r0688wztWrVKl1wwQUqLCxUz549dcIJJ2jlypWSpEceeUTXXntto/v+97//1WuvvabR\no0fr8ssvV0lJyVc3YjPy8vL04Ycfxn1/SRo9erROPPFEFRQU6He/+13k967r6sorr4wMN506dWrk\nb0899ZQuv/zyqPWXlpYqOzs7odoAAMnBcFMASHNt27aNDDctLy/X1VdfrVWrVun111/XFVdcoVAo\npFNOOUUXXHCBli9frjFjxuihhx7Seeedp9mzZ6t79+6SajtcdfPSevXqpR49ejTqoLzxxhuR4Y4X\nXnihtm3bpt27d0tq/qjUG2+8oYULF0qSRo0apRNPPDHyt9NPP13Dhg2LrHdkvStWrNDxxx8fNX/d\nkbEJEyZoypQpkqRXX31Vq1evjqyze/du7d27V0uXLtXf/vY3SdK3vvWtBrXU16pVK33/+9+XJF11\n1VUaN26cJOn666/XI488ol//+td66qmntGLFikb3fe655xQOh3X00Ufrkksu0YwZMzR79uwGndv6\nog1BbcnRwpa49dZb9cUXX6h3794Nnrep4aZvvfWWTj75ZJ166qk65ZRTVFBQoJ07dyojIyNydPqu\nu+7SunXr9MYbb3hSHwDAWxxJBABEnH322dq6dav+85//KBQKNRqeWNcheffdd3XyySdr48aNUR+v\nVavGu5l4Oi7N3ee4446Lul4oFFLr1q116NChyO+++OKLZp+nLp/rulq2bJkqKipUUVGh6urqyHPF\nWn/9f7dx48bpxRdf1HPPPaehQ4c22cksKSnRK6+8ou7du2vIkCHavn27Fi9eLEnq0KFDgxPXbN++\nXR07dmz2uSsqKjRgwICY6m1KKBRqsi2bUlJSotWrV6t79+7q1auXdu3apaeffjryOLNmzdJHH32k\nWbNm6ec//3nCtQEAvEcnEQAQ8eGHH+rQoUPq2LGjzjvvPD355JM6dOiQ/vOf/2jp0qUaNmyY1q1b\np1//+teqqKjQiy++qOXLl0uq7Qz99a9/leu6WrNmjT755BP17du3weOfd955+stf/iKpdq7gySef\nrPbt26t9+/aRI4pHOvfccyPzC19++WXt2LGjyfWOrPfvf/+7hg0bptNOO00ffPCB9u/fr507d+q1\n115rcL8nn3wy8v9zzjlHknTRRRdFTgAjKTJE9Pzzz9fjjz8uSXrxxRebreXQoUP661//Kkl6/PHH\ndd5550mqneM3atQo3XLLLfrBD37Q6H67du3S66+/rurqaq1du1Zr167VnDlzIkM2w+GwHn30UUlS\nTU2N/vKXv+gb3/hGkzUsWbJEDz30kG644YYm/+6FIzvMdblXrVoVqX/hwoUNhpzW3ee2225TdXV1\nIM9qCwC2Y7gpAKS5ujmJUu0H+OLiYoVCIX3ve9/Tm2++qZycHIVCIf3qV7/SKaecopEjR+r//u//\n1LlzZz388MMqKCjQihUrFAqFdNppp2nYsGHatWuX5s6dq6OPPrrBMMkZM2bo2muvVU5Ojo477rjI\nCWTGjBmjSy+9VIsWLdKcOXN07rnnRuqbPn26Jk6cqEcffVTDhw9X586d1b59e+3atavBUMvm6pWk\nyy+/XNnZ2erevXuj4ZE7duxQTk6OjjnmmEhn5sEHH9Stt96qnJwcHTx4UBdccIF+97vfRWopKSnR\nOeeco9NPP73Jf9PjjjtOy5cv17333qtOnTpFOqKSdMUVV+hvf/ubLrrookb3W7hwoUaMGKE2bdpE\nfjd27FjdcccdOnDggO655x7dcsstys3Nleu6Gj16tK666ipJtUfpnnzySb3++uv6/PPP1aNHDy1Y\nsKBBR73+v1dzw1eb0ty6V155pdq2bStJ6tixo+655x5lZmaqc+fOkXXOO+88rV69OjI/tP7j3H33\n3SosLFRpaWmL6gAApEbI9WrCAgAgrf3gBz/QmDFjIvPvvLJ//34dddRROuqoo/Tmm2/q1ltv1Tvv\nvOPJY3fv3l1vv/125IyuqTBr1izt3r1bhYWFKXtOAABiwZFEAIDR1q9fr8svv1yHDh3S0UcfrYce\nesizx071dRW/973vae3atY2GvAIAYBKOJAIAAI0bN05r165t8Ltf/vKXGjlypE8VAQD8QicRAAAA\nABDB2U0BAAAAABF0EgEAAAAAEXQSAQBpr6CgQPfcc4+vNbRq1UqffPKJrzUAACDRSQQA4CuvGeg4\njrp169bix4t1fQAATEInEQAASZzHDQCAWnQSAQBpp6KiQoMHD9bxxx+vCRMm6L///W+z6+7du1ej\nR4/Wpk2b1L59ex1//PHavHmz9u3bp8mTJ6tr167q2rWrpkyZov379ze7/vLlyzV8+HCdeOKJ6tKl\ni370ox/pwIEDKUwNAEDL0EkEAKSV/fv365JLLlF+fr527Nihyy67TPPnz292uOlxxx2n0tJSdenS\nRbt379auXbvUuXNn3XfffVq+fLlWrlyplStXavny5br33nubXb9169aaPXu2tm3bpjfffFOLFy/W\n7373uxSnBwDgq9FJBACklfLych08eFCTJk3SUUcdpfHjx+vMM8+Mep+mhqI+/vjj+p//+R917NhR\nHTt21PTp0/Xoo482u/7gwYM1bNgwtWrVSqeffrpuvPFGLVmyxJtQAAB4qLXfBQAAkEqbNm1S165d\nG/zu9NNPj3lO4qZNm3T66adHlk877TRt2rSp2fX/+c9/6sc//rHefvttff755zp48KCGDh0aW/EA\nAKQARxIBAGnl1FNP1caNGxv8bt26dVHPbtrU37p06aKqqqrI8vr169WlS5dm17/llls0YMAAffzx\nx/rss89033336dChQ3GmAAAgeegkAgDSyjnnnKPWrVvrwQcf1IEDB7RgwQKtWLEi6n06deqkbdu2\nadeuXZHfTZw4Uffee6+2bt2qrVu36uc//7muvvrqZtffs2eP2rdvr2OPPVYffvihfv/73ycnIAAA\nCUq4k3jttdeqU6dOGjhwYLPr3H777erdu7dycnJUUVGR6FMCABC3Nm3aaMGCBZo3b546dOigp556\nSuPHj496n379+mnixInq0aOHTjrpJG3evFl33323hg4dqkGDBmnQoEEaOnSo7r777mbXnzVrlh5/\n/HEdf/zxuvHGGzVhwoQGRxyjHckEACCVQm6CF4ZaunSp2rVrp2uuuUbvvfdeo7+/8MILmjNnjl54\n4QUtW7ZMkyZNUnl5eSJPCQAAAABIkoSPJJ533nk68cQTm/37M888o/z8fEnSWWedpZ07d2rLli2J\nPi0AAAAAIAmSPidx48aN6tatW2Q5MzNTGzZsSPbTAgAQk1/84hdq3759o9u3v/1tv0sDACClUnIJ\njCNHtDLvAgBgmrvuukt33XWX32UAAOC7pHcSu3btqurq6sjyhg0bGl2fSpJ69eqlNWvWJLscAAAA\nAEh7OTk5qqysbPJvSR9uOnbsWP35z3+WJJWXlysjI0OdOnVqtN6aNWvkum5Sb/n5+Ul/jlTdyGLm\njSzm3WzJQRZzb2Qx72ZLDrKYeyOLmTdbsqQqx8qVK5vtwyV8JHHixIlasmSJtm7dqm7duqmwsFAH\nDhyQJN1000361re+pRdeeEG9evXScccdp0ceeSTRpwQAAAAAJEnCncSSkpKvXGfOnDmJPo0nsrKy\n/C7BM2QxE1nMY0sOiSymIot5bMkhkcVUZDGTLVlMyJH04aYmCYfDfpfgGbKYiSzmsSWHRBZTkcU8\ntuSQyGIqspjJliwm5EirTiIAAAAAIDo6iQAAAACAiJDruu5Xr5Z8oVBIhpQCAAAAAFaL1v/iSCIA\nAAAAICKtOomO4/hdgmfIYiaymMeWHBJZTEUW89iSQyKLqchiJluymJAjrTqJAAAAAIDomJMIAAAA\nAGmGOYkAAAAAgBZJq06iCeN7vUIWM5HFPLbkkMhiKrKYx5YcEllMRRYz2ZLFhBxp1UkEAAAAAETH\nnEQAAAAASDPMSQQAAAAAtEhadRJNGN/rFbKYiSzmsSWHRBZTkcU8tuSQyGIqspjJliwm5EirTiIA\nAAAAIDrmJAIAAABAmmFOIgAAAACgRdKqk2jC+F6vkMVMZDGPLTkkspiKLOaxJYdEFlORxUy2ZDEh\nR1p1EgEAAAAA0TEnEQAAAADSDHMSAQAAAAAtkladRBPG93qFLGYii3lsySGRxVRkMY8tOSSymIos\nZrIliwk50qqTCAAAAACIjjmJAAAAAGAIx5HC4eQ/D3MSAQAAACAADBhtml6dRBPG93qFLGYii3ls\nySGRxVRkMY8tOSSymIosZrIlS1WV43cJau13AQAAAACQzhzn8BHE4mIpK6v253A4NUNPj8ScRAAA\nAAAwxIwZtbdkY04iAAAAAKBF0qqTaMs4ZYkspiKLeWzJIZHFVGQxjy05JLKYiixmsiVLRobjdwmJ\ndxJLS0vVr18/9e7dWzNnzmz0961bt+riiy9Wbm6usrOzNW/evESfEgAAAACslJvrdwUJzkmsqalR\n37599eqrr6pr164688wzVVJSov79+0fWmTFjhvbt26f7779fW7duVd++fbVlyxa1bt3wnDnMSQQA\nAACA1EjanMTly5erV69eysrKUps2bTRhwgQtWrSowTqnnnqqdu3aJUnatWuXOnTo0KiDCAAAAAAw\nQ0KdxI0bN6pbt26R5czMTG3cuLHBOjfccIPef/99denSRTk5OZo9e3YiT5kQW8YpS2QxFVnMY0sO\niSymIot5bMkhkcVUZDGTLVlMyJHQIb1QKPSV6/ziF79Qbm6uHMfRmjVrNHLkSK1cuVLt27dvtG5B\nQYGyvrwoSEZGhnJzcxX+8sIgdf9YiSxXVlZ6+nh+LldWVhpVD8u1y3VMqYftxa7lOqbUw+urdpn3\nY/OWeX2xnOzlOqbUw/ZSu2zL9lLH68cvKipSZWVlpL8VTUJzEsvLyzVjxgyVlpZKku6//361atVK\nU6dOjazzrW99Sz/72c907rnnSpJGjBihmTNnaujQoQ0LYU4iAAAAAKRE0uYkDh06VP/6179UVVWl\n/fv368knn9TYsWMbrNOvXz+9+uqrkqQtW7boo48+Uo8ePRJ5WgAAAABAkiTUSWzdurXmzJmjUaNG\nacCAAfr+97+v/v37a+7cuZo7d64k6a677tJbb72lnJwcffOb39Qvf/lLnXTSSZ4UH6sjD+EGGVnM\nRBbz2JJDIoupyGIeW3JIZDEVWcxkSxYTciR8mtHRo0dr9OjRDX530003RX7u2LGjnn322USfBgAA\nAACaVVkpfTn9DglKaE6il5iTCAAAACBeM2bU3tAySZuTCAAAAACwS1p1Ek0Y3+sVspiJLOaxJYdE\nFlORxTy25JDIYiqymMNxDh9BLCx0Ij8HOZYJbZLwnEQAAAAA8EM4fHgeYlUVw029wpxEAAAAAIHH\nnMTYMCcRAAAAgNU4s6l30qqTaML4Xq+QxUxkMY8tOSSymIos5rElh0QWU5HFVI7fBXjChDZJq04i\nAAAAACA65iQCAAAAQJphTiIAAAAAoEXSqpNowvher5DFTGQxjy05JLKYiizmsSWHRBZTkcVMtmQx\nIUdadRIBAAAAANExJxEAAAAA0gxzEgEAAAAALZJWnUQTxvd6hSxmIot5bMkhkcVUZDGPLTkkspiK\nLGayJYsJOdKqkwgAAAAAiI45iQAAAACQZpiTCAAAAABokbTqJJowvtcrZDETWcxjSw6JLKYii3ls\nySGRxVRkMZMtWUzIkVadRAAAAABAdMxJBAAAAIA0w5xEAAAAAECLpFUn0YTxvV4hi5nIYh5bckhk\nMRVZzGNLDokspiKLmWzJYkKOtOokAgAAAACiY04iAAAAAKQZ5iQCAAAAAFokrTqJJozv9QpZzEQW\n89iSQyKLqchiHltySGQxFVnMZEsWE3KkVScRAAAAABAdcxIBAAAAIM0kdU5iaWmp+vXrp969e2vm\nzJlNruM4jvLy8pSdna1wOJzoUwIAAAAAkiShTmJNTY1uu+02lZaW6oMPPlBJSYlWr17dYJ2dO3fq\n1ltv1bPPPqtVq1bp6aefTqjgRJgwvtcrZDETWcxjSw6JLKYii3lsySGRxVRkMZMtWYqKHL9LSKyT\nuHz5cvXq1UtZWVlq06aNJkyYoEWLFjVY5/HHH9f48eOVmZkpSerYsWMiTwkAAAAA1qqs9LuCBOck\nPv3003rppZf00EMPSZIee+wxLVu2TL/97W8j60yZMkUHDhzQ+++/r927d2vSpEm6+uqrGxfCnEQA\niMpxJEbsAwBgtxkzam/JFq3/1TrRB/4qBw4c0DvvvKPFixfr888/1/Dhw3X22Werd+/eiTw1AKQd\nOokAANjJcWpvklRYePj34bA/+/6EOoldu3ZVdXV1ZLm6ujoyrLROt27d1LFjR7Vt21Zt27bV+eef\nr5UrVzbZSSwoKFBWVpYkKSMjQ7m5uZET3dSNMU5kubKyUpMnT/bs8fxcLioq8vzfx6/l+uPHTagn\nkeUjM/ldD9uLXa+vqipJ4vVl2jLvx+Yt8/oyc9mW11f9DKbUk8jy009Xas4cthe/l2t/VbtcVSXN\nmFF/m/Hm+YqKilRZWRnpb0XlJuDAgQNujx493LVr17r79u1zc3Jy3A8++KDBOqtXr3ZHjBjhHjx4\n0N27d6+bnZ3tvv/++40eK8FSWqSsrCzpz5EqZDETWcwT9BxlZa47fXrtTSqL/BzwWIFvl/rIYh5b\ncrguWUxlU5b8/DK/S/CMLe2SqjaJ1v9K+DqJL774oiZPnqyamhpdd911mjZtmubOnStJuummmyRJ\ns2bN0iOPPKJWrVrphhtu0O23397ocZiTCADRpWqOAgAgfbBvMY/j6Msji8kVrf+VcCfRK3QSASA6\nduQAAC84Tu1Nqp3/Nn167c/hcGo6JzBDtP5XqxTX4qv648iDjixmIot5bMkhSRkZjt8leMamdiGL\neWzJIZHFVEHPEg4f/uIxP9+J/Bz0DmLQ26WOCTnSqpMIAEGWm+t3BQAAIB0w3BQAAABIU6ma/wbz\nMCcRAAAAABDBnMQvmTC+1ytkMRNZzGNLDokspiKLeWzJIUlFRY7fJXjGpnYhi5lsyWJCjrTqJAIA\nAARJZaXfFQBIRww3BQAAMBSXvgGQLNH6X61TXAsAAACiOPIadnW4hh2AVEmr4aYmjO/1ClnMRBbz\n2JJDIoupyGKeoOfgGnbmI4uZbMliQo606iQCAAAAAKJjTiIAAIChuIYdko3XWPriEhgAAAABxId3\nJJsBIxthoLTqJJowvtcrZDETWcxjSw6JLKYii3lsySGRxVQ2ZamqcvwuwTO2tIsJOTi7KQAAAJBG\n6p9Bt7hYysqq/Zkz6KIOcxIBAACANMW1ONMXcxIBAAAAAC2SVp1EE8b3eoUsZiKLeWzJIZHFVGQx\njy05JLKYyqYsGRmO3yV4xpZ2MSFHWnUSAQDwWmWl3xUAQPxyc/2uACZiTiIAAAlgPg8AIIiYkwgA\nAAAAaJG06iSaML7XK2QxE1nMY0sOiSwmcZzDRxALC53IzwGPFfh2qWNLDokspiKLmWzJYkIOrpMI\noEmVlVwrCWhO/WuJVVUx3BQAYBfmJAJoEvOsgJZhWwEABBFzEgEASBKOuAPpx4DRgEBSpVUn0YTx\nvV4hi5mCnsXGeVZBb5P6yGKmykrH7xI8Y0u72JJDIoup5s1z/C7BMza1iy1ZTMjBnEQAEfXnWS1c\nyBA6oCW4TiLQMsx1B4KDOYkAmhQOB/sIIpAqzEkEWibo24rjHN4vFhZK06fX/lz/C1YgSKL1vziS\nCCCi/g5wyZLDO3N2gEBDR35YrMO2AtjryO07SB3eUCgU1/04gOMPx/F/X8KcxIAii5mCniUcPvxN\nb37+4TmJfr9RJSLobVIfWcxh47YiBb9d6tiSQwp+FhvnuktSVZXjdwkxcV232ZtUFuVvwRL07aWO\nCXNeOZIIAACApKh/9K28PFhH36LJzfW7AiC5Ep6TWFpaqsmTJ6umpkbXX3+9pk6d2uR6K1as0PDh\nw/XUU09p3LhxjQthTiJgFBOGOgBBwLYCtAxz3c0UCkl8BPefH3NekzYnsaamRrfddpteffVVde3a\nVWeeeabGjh2r/v37N1pv6tSpuvjii+kIAgHBh16gZdhWkEw2fQmRleV3BYC5TJvzmtCcxOXLl6tX\nr17KyspSmzZtNGHCBC1atKjRer/97W916aWX6uSTT07k6RJmyzhliSymIot5bMkhkcVUZDGPLTkk\nM+YmJaL+nMTiYnvmJNr0GsvPd/wuwTO2tIsJc14TOpK4ceNGdevWLbKcmZmpZcuWNVpn0aJFeu21\n17RixYq4z64EAACAYKl/dKSqyv+jI2isoMDvCnAkE+a8JjQncf78+SotLdVDDz0kSXrssce0bNky\n/fa3v42sc9lll+knP/mJzjrrLBUUFGjMmDEaP35840KYkwgAAGDt9fiCfp1EwDZJm5PYtWtXVVdX\nR5arq6uVmZnZYJ23335bEyZMkCRt3bpVL774otq0aaOxY8c2eryCggJlfTlgPSMjQ7m5uQp/+W5Y\nd/iYZZZZZplllllm2fblGTNql6uqHIXD/tfjxXI4bFY9LLOcbstFRUWqrKyM9LeichNw4MABt0eP\nHu7atWvdffv2uTk5Oe4HH3zQ7PoFBQXu/Pnzm/xbgqW0SFlZWdKfI1XIYiaymMeWHK5LFlORxTy2\n5HBd183PL/O7BM/Y1C5kMVPQskiK6+bl8zcnoSOJrVu31pw5czRq1CjV1NTouuuuU//+/TV37lxJ\n0k033ZTIwwMAAKQ1E+YmAUgOt5mhnqGQI9cNp7aYI2twm6suxZiTCAAAAKQWc0XNk6prV0brf9FJ\nBAAAMJTjSF9OJwKSIlUdErScCZ3EVsl/enPUTd60AVnMRBbz2JJDIoupyGIeW3JIwb9OYn02tUtR\nkeN3CR5y/C7AM7a8xky4dmVadRIBAACARFVW+l0BbGbCtSsZbgqgSQxxAgB/OE7tTbLrOok2sWke\nH8NN01fSrpMIwF50EgHAH/U7g45jT2ck6I7svNeh8w4bpdVwU1vGKUtkMZVNWaqqHL9L8IRNbUIW\nM5HFPLbkkKSdOx2/S/BM0NslHD58BDE/34n8HPQOognz37wS9NdYHRNycCQRQET9b0mLi6WsrNqf\n+ZYUAFKn/nvxypWHjyTyXmyOzZv9rsA7Jsx/g3mYkwigSTbNtwASFQqF4rof+zUkivdiMxUUSPPm\n+V0FbJWq7Z5LYAAAkADXdZu9SdH+BsBGdSNtgGSoP+fVL2nVSTRhfK9XyGImm7JkZDh+l+AJm9qE\nLKZy/C7AM7a0iy05JHvei6Xgt0vdSYRmzJAKCw/PSQx4rMC3S332ZHH8LoA5iQCalpvrdwUAAJve\niysrgz3cqy3iAAAgAElEQVSnsv6c0KoqhgHDbsxJRMrEM6eH14R/uAQG0DLMGQNaxqZthSxIplRd\nu5I5iTAC83mChQn5QMvw4QpIPzZ9iWrC/DeYJ606ifaMU7YrC9fnMdPrrzt+l+AJm9qELGYii3ls\nySEFP4ut8/hMmDPmHcfvAjwT9O2ljgmfjZmTCN9xfR5z1L8215o1XJsLAJAY5vEBsTPhszFzEgEP\nBX0eX/1OYmGhNH167c90ElOLa/IBsBFz38yUqvlvMA9zEoEUsWSUA3wW7Zp80a/XBwDm4stGIDjS\nqpNoyzhliSymqqpy/C4hIeHw4W96e/Y8PHckyDt2m15fZDFTQYHjdwmesaVdbMkhBTNLKBRq8nbh\nhU3/Pt7RE34KYrs0x4T5b16xpV1MyMGcRCBB9YdoFhdLWVm1Pwd9iGavXn5XAARDcTFnA0byBPHa\ngs2NbAiFHLluOLXF4CuZMP8N5mFOInxn0xwFm7IEfX4lkCrM50Ey2bRfsSkLkEyp2laYkwijcX0e\n/zQ37Me2YUEAAP/RQQRaxoTPxmnVSTRhfK9XbMpi0/V5MjIcv0uISbSToJSVlVlxghSbthXmvpnK\n8bsAz9jSLkHPYeu1BYPeLvUVFTl+l+AZm9rFniyO3wUwJxHwUm6u3xXAZsx9A9ID1xY0X2Wl3xUA\nycWcRPiO+TxmYu6IedhWzMS2gmTi9WUmm9rFpiy2SNX+Plr/i04ifMcHXzPZ0i42nYDHljYB0HI2\nvYcFXf2zmRcWStOn1/4c9LOZs28xjwmdROYkBpRNWbg+j6kcvwvwxLx5jt8leMjxuwDP2LStkMU8\ntuSo5fhdgGeCPq+6/rWE8/PtuJZwLcfvAjxjy7ZvwmfjtOokwkxcnwcAzMA8KyRTcbHfFXhn82a/\nK4DNTPhszHBTAE0K8vATW4cEMW8EycZrDMkU5P3KkS6+WCot9bsKb9jULohNtP4XZzcFYJ0jO4O2\nfOi1JYfEPCsAwfbf//pdAZrCvsU7CQ83LS0tVb9+/dS7d2/NnDmz0d//8pe/KCcnR4MGDdK5556r\nd999N9GnjJst45QlspjKpiwmjIf3QlWV43cJnrHp9WXTXNGgz7Oy8Zp8Nm0rNmUJ+ty3+tvKkiVm\nbisnnVR7ZDCWm+TEtP5JJ/mdsnm27FtM2O4TOpJYU1Oj2267Ta+++qq6du2qM888U2PHjlX//v0j\n6/To0UN///vfdcIJJ6i0tFQ33nijysvLEy4cQHKZMB7eC1y7EskW9OtXck0+s1VWcmQELbdjR+xD\nR2M9+lbbsYTtEuokLl++XL169VJWVpYkacKECVq0aFGDTuLw4cMjP5911lnasGFDIk+ZkLBV77Jh\nvwvwjOOErdkB2vQasyXL5Mlhv0vwTNDbpP5c0eLisL7cdQR+rqhN78dZWWG/S/BE0LeV+nbuDPtd\ngmemTw/7XUJC6r9XFRWFrflCJejbi437FhM+GyfUSdy4caO6desWWc7MzNSyZcuaXf/hhx/Wt771\nrUSeEl+yacx1YSHfXAPpwNa5ojaxZb8CMwV9m6/fGfnss8N5gtwZsYGN+xYTPhsnNCcxFMPx5rKy\nMv3pT39qct5iqpgwvtcrNs2zCvochfpseo3ZkqWoyPG7BM8Efe5bfbyHmcrxuwBPBP39y8Z5olLw\n26Uhx+8CPGNTu5SXO36X4BHH7wISO5LYtWtXVVdXR5arq6uVmZnZaL13331XN9xwg0pLS3XiiSc2\n+3gFBQWRoasZGRnKzc2NHAKvewEnslxZWenp46V6ubLy8LCT4uLai1llZdUdjva/PpYPM6WedN5e\n6pYrK82qJ5HlurlvptSTyHK7dpWqG6ZpQj2JLEuVqh3dYUY96bQcy5fV9ZWVlRlRf7TlGTNql8vL\nKxUO+18Py2HV/qp2eeHCuo577bIp72eSo1jfj+LZ35uS98jlqio73o/reP34RUVFqqysjPS3okno\nOokHDx5U3759tXjxYnXp0kXDhg1TSUlJgzmJ69ev1ze+8Q099thjOvvss5svhOskxsSma1lxfR4z\n2fIasyWHxLaSCiedVHvih2Q78URp+/bkPw+Cz6b3MJsUFJh5wqpU7CdM3heFw9IR/axAStW/cdKu\nk9i6dWvNmTNHo0aNUk1Nja677jr1799fc+fOlSTddNNN+vnPf64dO3bolltukSS1adNGy5cvT+Rp\nAWPVfnvldxXeMGE8fLwc5/BOorDw8O9rv41PfT0IjnjODBgPzg6IluI9y0y2nAHcBvX3+UuWMFfU\nK60SfYDRo0fro48+0scff6xp06ZJqu0c3nTTTZKkP/7xj9q2bZsqKipUUVHhawfxyEO4QZaR4fhd\ngmdsuR6fZM/1eWo5fhcQt3D48Lfv+fmH5/MEf2fh+F2AZ2x6PyaLeWzJUcvxuwDP2DSvurLS8bsE\nz7C9JF+s16+UYrt2ZTKuX5nQkUTACzZ9G7d5s98VAABgpqBfU7S+ykq/K0Cd+kcM6+aKmibWUSrx\njEzzeoRKQnMSvcScxNgwR8EcRw5tnD699uegD3Mwec5BLEwdAszcNzOlbh6IHduXqdhHmsmm1z1z\nEpP7HPGiXWK9T/P9LzqJAcUO0Ey2TJiWzN4J2IDOiJloFzvY9O9r6hdd8Qh6uwThS2FTOyOpYur2\nYmq7ROt/JTwnMUiCPubacbhukonqt8uSJWa2S6xj4eMZD+/1WHiv2HSdxKBvK/WRxUz2ZHH8LsAz\nzHU3R/357jk59sx3t2e7l4L+GqtjQpswJzFA6n9TVV7OkURTBKFd4jljY6zfxpl6tkbmjQAAvFD/\nSOLKlZxFE3ZjuGlA2TSs0aahs4yFT+5zxMPU11e6D2tM5yFBqXyedBX0f98gDGtM93nV7O+T+xzx\nKiqSJk/2u4rGTG2XpF0nEf455hi/K/BOkK/HBzNxnUTzmdpJBExQ/72qqsrMfWS6X1M0K8vvCtCU\nhQvN7CQGEXMSA6T+3LeXXjJz7lt8HL8L8JDjdwGeCfL2Yut1EoPcJkcqL3f8LsEzNrWLLVlsuv4u\n1+Mzk03Xq7apXXbudPwuwRMmtAlHEgMkCNeBSUf1j1oVFx/+dpGjVkBD9beVl15iPo+JKivtaAub\nrr/bubPfFaApubl+V4A6RUW1n4ul2rmide9hl1zCUcVEMCcxQIIwRyEeJo9tj1U6z38ztR1NHdaY\n7nPfcnPNPKlQureLqfOs0k0Q9vfpvq2YKp3395K55+wwtV2Yk2gJjiQCsTPlA9WRXIWkFMy1cev9\n129BODNgOrZLfVVVflcAqfE2wf4eaJmdO/2uwB7MSQwsx+8CmpSK6/GZfE0+U+co1H7wje3mxLi+\nm4pP1nEwdbsPya39yi+Gm1NWFvN9QgZ1ROrPFT3xRDPniqZjuwThWq+xMnW7j0dVleN3CZ6xqV3I\nYqajj3b8LsETJrQJRxIDJAjfwqfienySuWc7M3WOQkhu7MMcYmyYUMjE4yIwSf33sB07zHwPA0xj\n6n4FZkrFaAhTR0JI5h5ECCLmJAZUOs99S+Xz2MLUsfDePn98e0W/3nfScVthnlXqnydWGRl2DNcy\ndR8ZDe9h/j6PLdJhf38k9i3xP0e0/hedxIAydQeYDjuNoO3IJXPfnNJZOmwr0aTzyQVS+TwtEYQP\nWLEy6d83Uezvg/cFJPt7/5i6b0nZELgYGyZa/4s5iYYLhUJN3goLc5v9W9AErV1c1232VlZW1uzf\ngiZo7dIcW3JIdmU55hjH7xI8Y1O7mDrfPXaO3wV4Jhx2/C7BM0HcVprbp+fns783RRDmVcc6392E\nue7MSTRcc282juMoHNSveAGkjea/uMpRKLSyyb8E8UNWkHHmbMQj3c8EXFzM5WJMUf89rLyc9zCv\nMNwUnkqX4SdBk+7DT0zEtmKmdG8XW66TaOq/r03SfVtJ57r8zh591Nx0SYVN/oVhwE3dh+Gm1uFb\nEjPRLgBM19xUhVAopOLisDVTGQDYKfq0nxnWDAP2W1p1EoM+5rq+wkLH7xI8Q7ukRqzXogyFYrt+\n5Ykn+p2waQUFjt8leMambYV28U8QP2DFeg1em66/G7TXVzQ2ZbFp3ivtYh4T2iStOolAuopx7nNk\nuEIs62/f7m/G5hQX+10BmkK7IBZ11+Bt6S2Ocz5oxw6/UzbNhuG/QKqwvXiHOYkB5fd48OYwR8HM\nuuJhSxZTc7CtpHddpuZP58stmNom6V5XKp7npJNS8yXBiScm/wvVdN5WJHNrM7VduE6ihdJ5I0jl\n88TK1LriYUsWU3OwrZhbVyqk4sNiPExuFxM/YKWCyXWlgi0dq1Q9TyraxdT3L8ns7cXE9zBOXPMl\nE8b3esfxuwDP0C6mcvwuwCOO3wV4hm0l+eIbmu3EfB9TP2CZ2i6xYltJvnTfVkx9jaWiXUxtk1qO\n3wV4woTXF9dJNEC8wxxi/bbI5G9+TJSKdqFN0luqvvEFAACIBcNNDWDbMAeymPUc8TJ1blIq5o6Y\n2nk39fVi03yeeJjaLvEwNUs6vx+bWlc8TM1i02eXeJhaVzxMzWLqexhzEg1n05sTWcx7Dtukc7uk\ne13pnj8VTM1i03bPF11+V9EY72Fm1mXTtmLqXFHmJH7JhPG9XjE1i6sYL0wVCsmJ9WJWoVDt8xjI\n1HaJhy1ZbMlRy/G7AM/QLskX67UFaz/EODGtz7UFYxf75Txin8dn6uU8TN1W4mHyayx2jt8FNCnW\nbSWe7SVV20qy54kmY64ocxINUNuxSsXzHP5vsoTkxv5tlONI4XBszxNKdpLUtEsq2gRINpvew+KR\nn+93BU2r+4AVi1jfjlN1hkvYwdRthfcwvyuAiRhuagCbhjmQxbznsE6qPpUa2DCmvl5s2u5tYtV7\nmEXbvVXtYgnew8yUzttK6l6TSRxuWlpaqn79+ql3796aOXNmk+vcfvvt6t27t3JyclRRUZHoUwJI\nYyHFOP4ijlvIwG96JWn6dL8rAPyRzts9APghoU5iTU2NbrvtNpWWluqDDz5QSUmJVq9e3WCdF154\nQR9//LH+9a9/6Q9/+INuueWWhApOhE3jx8liJpuyFBQ4fpfgCZvaJBx2/C7BMza1i6lZUjFHnPnh\nyUeW1Ij99Aixzd8Nhcy9JJGp7ZLO72EmzBNNaE7i8uXL1atXL2VlZUmSJkyYoEWLFql///6RdZ55\n5hnlfznY+ayzztLOnTu1ZcsWderUKZGntk4qRtKY+uZksmS3i8ltUlwszZvndxUICtvfw0JxBvRz\nGkUq5oinYn64bZjvbp54NlNThynaJJ3fw0yYJ5pQJ3Hjxo3q1q1bZDkzM1PLli37ynU2bNjgSycx\nHOPJUVIlvjensDVvTrSLqcJ+F+AJU19f8TA1SzpsK+kyZ97U11isTM4R6wffcDzPYegHX5PbJXZh\nvwvwjMntEvv3c+GY1jb1y/p588J+l5DYcNOWfrN65M413m9kAUCKZ1iQHUOCgFixrcAkM2b4XQGa\nYmq7xDO9ONb7mXg9UVMkdCSxa9euqq6ujixXV1crMzMz6jobNmxQ165dm3y8goKCyNDVjIwM5ebm\nRr7dqBsvnchyZWWlJk+e7Nnj+bksFclxvP338WK57hucWO5ffyx8y/M7X44o8Ddv4/ySFDamnsSW\nKyWZt724bmzrh0KOysoU8/OZ+vqqv834XY+Nr694louKijzfX3mx7Lqx3z/e7SUVy7HsX+LZ38ez\n/4pnWXK+PDpyeLn+8zdeLpKUG8P6jtq1S12eWJYLCx3V/TOYUE8iy1+mMKaeRJYLCys1Y4Yd78em\nfj6O5/UVTsL+vqioSJWVlZH+VjQJXQLj4MGD6tu3rxYvXqwuXbpo2LBhKikpaTAn8YUXXtCcOXP0\nwgsvqLy8XJMnT1Z5eXnjQkLJvwSG4zj1XkTBFgo5kQ8AJolnjH487WLqXABT2yUetmSxJYdUezIh\nE4ageMGmdmHfknyxvuezXzETWcxEFvOkar8Srf+V8HUSX3zxRU2ePFk1NTW67rrrNG3aNM2dO1eS\ndNNNN0lS5Ayoxx13nB555BENHjw4piLRmLk7s/S+1pCpdcVjxgxzh6DEwqY2IQuSzdR2SUVd6Zw9\nVWzKYss+UrKrXWzKkgpJ7SR6hU5ibEx9c0r3TqKp7ZLOTH2txMOmLGwrZjL1NZaqs+eaOD/J1DaJ\nh01ZbGJTu9iSJVX7yGj9r1bJf3pzNBxHHmxcL81MtIuJHL8L8JDjdwExC4VCTd4KC5v+fRBPbGbP\ntiKZ+hqL/eQVTsz3MbGDWMvxuwAPOX4X4Bm2ezPl5zt+l+CJwkLH7xISO3EN0BTbr5eGYDHhWkPp\nrLlvKG2axxc0X9URb+7PjPZBong/NpNN7ZKb63cF9mC4KXxny9AASXIcyZbPvUVF0pcnB4QhbNpW\ngGSyaVthaDbQUPQvuqZLKmzyL0HqZ6Ru+hbDTa1j1SiHgGluiFwoFNKFF84L1BC6aFmmTHEClSUd\nTJ/udwWwHfsW89BBBBpyXbfZ2/TpM5r9G2KTVp3EII4fb74zUmDRB3jH7wJiEu3NKScnK1BvTtGz\nNP/3ILFpu482jy9o235RkeN3CZ4J4musOQ884PhdgkccvwvwjE2vL5uyFBQ4fpfgmaC3i+McPuJe\nWOhEfg5CrOb35/7v75mTaLjmPpDXXi9tXmqLQZMc5/Ab0cqVh7/1DYeDN/S0qEhauLD255UrD9d/\nySUMPU2ldJjHV1npdwVoyubNflfgDZvmWMFf0T6QFxc3/XsTv0yNt2NhYpYj1f+8VV4erKPvJu/v\nmZMYIPU7I4WFh4eeBbEzUl/Q547Y2i7hcDC+hatj8w7QRsyzMoet72EA0k/QPrv4LVr/iyOJAVJ/\nh113aN0GQZ9nZWu7BA2dPfMd2RmpQ2cEsEs8X9rxHg4vZGX5XYE9mJMYUDt3On6X4Bmbri14zDGO\n3yV4Jjvb8bsET9i03Qc9Szh8+Ahifv7heSNB7yDa1C6dOtnRLkFvk/qCOPetufnsZWVlVsx1l+x6\njQU9S/05icXFwZqT2BwT2oQjiQFi09w3W118sd8VeOfSS/2uAEAq1N+3bNnCvsU0xcUSpyAAmlf/\nvaqqihFdXmFOYkAxnwdAkNl0TVGbFBTQIfEL86qBxPH5ODZcJ9FCVVV+VwAA8aODaCZb5vME8UNi\n80M0m79UER1EoCH2Ld5Jq06iCeN7vVJe7vhdgmdsaheymMeWHBJZTGVTlq1bHb9L8ERhoeN3CZ6Z\nN8/xuwTP2LStkMVUjt8FeMKENkmrTqJNdu70uwLvMLTJTFzHDkg/H3/sdwUAABMwJzFAbL2WVdCv\nk2grxvUD6ceWa4wFfb9i6/4egFm4TqIlKisb7rzrfs7IYKcBAIhP/Q7JkiXBOrtptJO9RDsPjOlf\nSh/5b88XdgBSLa2Gm5owvjcRkycf3pkfd5wT+XnyZH/raqlQKNTkTWr69/Ge6c1PQX+N1b/WUGEh\n1xoyDVnMZFOWoM3nied6fKZ3EI9UVeX4XYJnbNpWyGImW7KYkIMjiQFS/9vevXuD9W2v1Pw3t47j\nKByEAGmAaw0B6af+dj9rFtu9aXJz/a4AQDpiTmJAXXyxVFrqdxWwGXMSzcO1BZEMzH8DgPTEdRIt\ndPbZflcA2/Hh0DwGjD4BAABpIK06iSaM7/VKRobjdwmesaldioocv0vwkON3AZ6w6fXF3CQzBT1L\nOHx45EBOzuG5yEH+oijobVIfWcxEFjPZksWEHMxJDCjmKJiJawvCa/WHAhYXS1lZtT8zFBDJ0Lmz\n3xUAAEzAnETAQ8zjQzLx+kKyMe8VANIH10kEkujIkz7U4UgPgKDhPQsAIDEnMbDIYo7683ny8+2Y\nzyMFv13q2JJDYi6yqchiHltySGQxFVnMZEsWE3KkVScRAIKMucgAACAVmJMIeIj5PAAAAAiCaP0v\nOokAAAAAkGai9b/SaripCeN7vUIWM5HFPLbkkMhiKrKYx5YcEllMRRYz2ZLFhBwJdRK3b9+ukSNH\nqk+fPrrooou0c+fORutUV1frwgsv1BlnnKHs7Gw9+OCDiTwlAKQtrsMJtAzbCgAkJqHhpnfccYc6\nduyoO+64QzNnztSOHTv0wAMPNFhn8+bN2rx5s3Jzc7Vnzx4NGTJECxcuVP/+/RsWwnBTAIiK6yQC\nLcO2AgBfLWnDTZ955hnl5+dLkvLz87Vw4cJG63Tu3Fm5X56Sr127durfv782bdqUyNMCAAAAAJIk\noU7ili1b1KlTJ0lSp06dtGXLlqjrV1VVqaKiQmeddVYiTxs3E8b3eoUsZiKLeYKew3EOHxUpLDx8\nHc6Axwp8u9RHFjOwrZiPLGYii3lMyNH6q1YYOXKkNm/e3Oj39913X4PlUCikUCjU7OPs2bNHl156\nqWbPnq127drFUSoApJ9w+PBlVaqqGEIHNIdtBQC885WdxFdeeaXZv3Xq1EmbN29W586d9emnn+qU\nU05pcr0DBw5o/Pjxuuqqq3TJJZc0+3gFBQXKysqSJGVkZCg3N1fhL9/x63rUiS7X8erx/Fqu+50p\n9SSyHA6HjaqHZbu2F5teX1lZZtXD6yvcIIMp9bC91C7XMaUeXl92vb5sW65jSj1sL8lZLioqUmVl\nZaS/FU3CJ67p0KGDpk6dqgceeEA7d+5sdOIa13WVn5+vDh066De/+U3zhXDiGgCIqqhImjzZ7yoA\n8zmOVO8zIwCgCUk7cc2dd96pV155RX369NFrr72mO++8U5K0adMmffvb35YkvfHGG3rsscdUVlam\nvLw85eXlqbS0NJGnjduR35YEGVnMRBbz2JJDkiorHb9L8IxN7UIWEzl+F+AZe9qELKYii3lMyPGV\nw02jOemkk/Tqq682+n2XLl30/PPPS5K+/vWv69ChQ4k8DQAAAAAgRRIabuolhpsCQGOOU3uTpMJC\nafr02p/DYYbTAQCA+EXrfyV0JBEAkFxHdgY5YyMAAEi2hOYkBo0J43u9QhYzkcU8tuSQpKoqx+8S\nPGNTu5DFPLbkkMhiKrKYyZYsJuRIq04iAARZbq7fFQAAgHTAnEQAAAAASDNJuwQGAAAAAMAuadVJ\nNGF8r1fIYiaymMeWHBJZTEUW89iSQyKLqchiJluymJAjrTqJAAAAAIDomJMIAAAAAGmGOYkAAAAA\ngBZJq06iCeN7vUIWM5HFPLbkkMhiKrKYx5YcEllMRRYz2ZLFhBxp1UkEAAAAAETHnEQAAAAASDPM\nSQQAAAAAtEhadRJNGN/rFbKYiSzmsSWHRBZTkcU8tuSQyGIqspjJliwm5EirTiIAAAAAIDrmJAIA\nAABAmmFOIgAAAACgRdKqk2jC+F6vkMVMZDGPLTkkspiKLOaxJYdEFlORxUy2ZDEhR1p1EgEAAAAA\n0TEnEQAAAADSDHMSAQAAAAAtkladRBPG93qFLGYii3lsySGRxVRkMY8tOSSymIosZrIliwk50qqT\nCAAAAACIjjmJAAAAAJBmmJMIAAAAAGiRtOokmjC+1ytkMRNZzGNLDokspiKLeWzJIZHFVGQxky1Z\nTMiRVp1EAAAAAEB0zEkEAAAAgDTDnEQAAAAAQIvE3Uncvn27Ro4cqT59+uiiiy7Szp07m123pqZG\neXl5GjNmTLxP5wkTxvd6hSxmIot5bMkhkcVUZDGPLTkkspiKLGayJYsJOeLuJD7wwAMaOXKk/vnP\nf2rEiBF64IEHml139uzZGjBggEKhULxP54nKykpfn99LZDETWcxjSw6JLKYii3lsySGRxVRkMZMt\nWUzIEXcn8ZlnnlF+fr4kKT8/XwsXLmxyvQ0bNuiFF17Q9ddf7/ucw2hHO4OGLGYii3lsySGRxVRk\nMY8tOSSymIosZrIliwk54u4kbtmyRZ06dZIkderUSVu2bGlyvSlTpuhXv/qVWrVi+iMAAAAAmK51\ntD+OHDlSmzdvbvT7++67r8FyKBRqcijpc889p1NOOUV5eXlGjK2tqqryuwTPkMVMZDGPLTkkspiK\nLOaxJYdEFlORxUy2ZDEihxunvn37up9++qnruq67adMmt2/fvo3WmTZtmpuZmelmZWW5nTt3do89\n9lj36quvbvLxcnJyXEncuHHjxo0bN27cuHHjxi3Jt5ycnGb7enFfJ/GOO+5Qhw4dNHXqVD3wwAPa\nuXNn1JPXLFmyRLNmzdKzzz4bz9MBAAAAAFIg7omCd955p1555RX16dNHr732mu68805J0qZNm/Tt\nb3+7yfv4fXZTAAAAAEB0cR9JBAAAAADYJ9CnHL322mvVqVMnDRw4MPK7lStXavjw4Ro0aJDGjh2r\n3bt3S5L279+vH/zgBxo0aJByc3O1ZMmSyH0uvvhi5ebm6owzztB1112nAwcOBDZLnbFjxzZ4rFTy\nKks4HFa/fv2Ul5envLw8bd26NbBZ9u/frxtvvFF9+/ZV//79tWDBgkBm2b17d6Q98vLydPLJJ2vK\nlCmBzCJJjzzyiAYOHKicnByNHj1a27ZtC2SOJ598Ujk5OcrOzo6M6ki16upqXXjhhTrjjDOUnZ2t\nBx98UJK0fft2jRw5Un369NFFF13U4LTe999/v3r37q1+/frp5Zdfjvz+7bff1sCBA9W7d29NmjQp\n0Fl+9rOf6bTTTlP79u1TnkPyLssXX3yhb3/72+rfv7+ys7M1bdq0wGaR/N3ve5mjjl/7fC+z+L3P\n9zKL3/t8r7KYsM/3sl383Od7mSNl+/x4T1xjgr///e/uO++842ZnZ0d+N3ToUPfvf/+767qu+6c/\n/cm95557XNd13Tlz5rjXXnut67qu++9//9sdMmSIe+jQIdd1XXf37t2R+48fP9599NFHUxUhwqss\nruu68+fPd6+44gp34MCBKUxwWKJZ6oTDYfftt99OYeWNeZXlf/7nfyLrua7rbt26NRXlN+Dla6zO\nkCFD3KVLl6ag+oa8aJd9+/a5J510krtt2zbXdV33jjvucGfMmJHKGJ7k2Lp1q3vaaadFXlP5+fnu\n4mBpu2UAAA1rSURBVMWLUxnDdV3X/fTTT92KigrXdWvfU/v06eN+8MEH7k9/+lN35syZruu67gMP\nPOBOnTrVdV3Xff/9992cnBx3//797tq1a92ePXtGXmNnnnmmu2zZMtd1XXf06NHuiy++GNgsy5Yt\ncz/99FO3Xbt2Kc3gdZbPP//cdRzHdV3X3b9/v3veeecFul383O97kaOmpibyeH7u871sE7/3+V5m\n8Xuf7/VrrI4f+3yv2sXvfb5XOVK5zw90J9F1XXft2rUNPmCdcMIJkZ/Xr1/vDhgwwHVd17311lsb\n7ARGjBjhLl++vMFj7d+/3x0zZkzKd3x1vMiye/du9+tf/7r7wQcfNHisVEsky4oVK1zXrd1hvPXW\nWymquHleZOnWrZv7+eefp6ji5nm5vXz00Udut27dklxx8xJtl5qaGrdnz57uunXr3EOHDrk333yz\n+9BDD6UuwJcSbZPly5e7I0aMiPz+z3/+s/vDH/4wBZVH993vftd95ZVX3L59+7qbN292Xbd2J1l3\nJuxf/OIX7gMPPBBZf9SoUe6bb77pbtq0ye3Xr1/k9yUlJe5NN92U2uKPEG+W+vzqJB7Jiyyu67qT\nJk1y//jHP6am6GZ4kcXv/b7rJpbDlH1+nUSymLLPrxNPlvLyctd1zdnn1/FiW/F7n18n3nYxZZ9f\nJ942SeU+P9DDTZtyxhlnaNGiRZKkv/71r6qurpYk5eTk6JlnnlFNTY3Wrl2rt99+Wxs2bIjcb9So\nUerUqZPatm2riy++2JfajxRPlnvuuUc/+clPdOyxx/pWd1NiyVL3N0nKz89XXl6e7r33Xl/qbkqs\nWeqGDtx9990aMmSILr/8cv373//2rf764t1eJOmJJ57QhAkTUl5zc2Jtl1atWmn27NnKzs5W165d\ntXr1al177bV+RpAUe5v07t1bH330kdatW6eDBw9q4cKFDbYhP1RVVamiokJnnXWWtmzZok6dOkmS\nOnXqpC1btkiqPclZZmZm5D6ZmZnauHFjo9937dpVGzduTG2AehLJYhqvsuzcuVPPPvusRowYkbri\nj+BFFhP2+/Hm2LRpkySz9vmJZpHM2ecn8voybZ/v1XZvwj4/3iwbNmwwap+fyLaSyn2+dZ3EP/3p\nT/rd736noUOHas+ePTr66KMl1c79yczM1NChQzVlyhSdc845OuqooyL3e+mll/Tpp59q3759Ki4u\n9qv8BmLNUllZqU8++UTf/e535Rp2PqJ42uUvf/mLVq1apaVLl2rp0qV69NFH/YwQEWuWgwcPasOG\nDTr33HP19ttva/jw4frJT37ic4pa8W4vUu2Y+IkTJ/pRdpNizbJr1y7dfvvtWrlypTZt2qSBAwfq\n/vvv9zlF7DkyMjL0+9//Xt///vd1/vnnq3v37o3aKpX27Nmj8ePHa/bs2Y3m4YVCoUCd5TqRLKbl\n9CrLwYMHNXHiRE2aNElZWVnJKjcqr7L4vd9PJIfrukbt873Y7k3Z5yeaxaR9vpfvYX7v8xPNYso+\nP9HXVyr3+a2T8qg+6tu3r1566SVJ0j//+U89//zzkqSjjjpKv/71ryPrnXvuuerTp0+D+37ta1/T\n+PHjtWzZMuXn56eu6GbEmsVxHL311lvq3r27Dh48qH//+9/6xje+oddee82X+uuLp126dOkiSWrX\nrp2uuOIKLV++XFdffXWKK28s1iwdOnTQscceq3HjxkmSLr30Uj388MOpL7wJ8W4vK1eu1MGDB5WX\nl5fagqOINcvq1avVvXt3de/eXZJ02WWXaebMmakv/AjxtMl3vvMdfec735Ek/eEPf1Dr1v68tR84\ncEDjx4/X1VdfrUsuuURS7TejmzdvVufOnfXpp5/qlFNOkVR7hLD+t58bNmxQZmamunbt2uCo9YYN\nG9S1a9fUBlHiWfyouTleZqk7Gcftt9+e2hBf8rpd/Nrve7GtlJeXG7HP96pNTNjne5HFlH2+l9uK\n3/t8L7KYsM/3qk1Stc+37kjif/7zH0nSoUOHdO+99+qWW26RVHtmtr1790qSXnnlFbVp00b9+vXT\n3r179emnn0qq/fbnueeeM+aDb6xZbr75Zm3cuFFr167V66+/HrmGpQlizVJTUxM5s9mBAwf07LPP\n+na21iPFmiUUCmnMmDEqKyuTJC1evFhnnHGGP8UfIdYsdUpKSnTFFVekvuAoYs3So0cPffjhh5HX\n2SuvvKIBAwb4U3w98bRJ3VCmHTt26Pe//72uv/76lNftuq6uu+46DRgwQJMnT478fuzYsZGjNMXF\nxZEd49ixY/XEE09o//79Wrt2rf71r39p2LBh6ty5s44//ngtW7ZMruvq0UcfjdwnaFlM4GWWu+++\nW7t27dJvfvOb1AeRd1n83u97lcOEfb5XWUzY53uVxYR9vtfvYX7u873K4vc+38s2Sdk+PykzHVNk\nwoQJ7qmnnuq2adPGzczMdB9++GF39uzZbp8+fdw+ffq406ZNi6y7du1at2/fvm7//v3dkSNHuuvX\nr3dd13W3bNninnnmme6gQYPcgQMHuj/5yU+aPItjELLUt3btWt/ObupFlj179rhDhgxxBw0a5J5x\nxhnu5MmTA90u69atc88//3x30KBB7je/+U23uro6sFlc13V79OjhfvTRR6mOEOFVluLiYjc7O9sd\nNGiQO3bsWHf79u2BzDFx4kR3wIAB7oABA9wnn3wypRnqLF261A2FQm5OTo6bm5vr5ubmui+++KK7\nbds2d8SIEW7v3r3dkSNHujt27Ijc57777nN79uzp9u3b1y0tLY38/q233nKzs7Pdnj17uj/60Y8C\nneWnP/2pm5mZ6R511FFuZmamW1hYGMgs1dXVbigUcgcMGBB5nIcffjiQWfze73v5+qrj1z7fqywm\n7PO9bBe/9/lev8b83Od7mcXPfb6XOVK1zw+5rmGT1wAAAAAAvrFuuCkAAAAAIH50EgEAAAAAEXQS\nAQAAAAARdBIBAAAAABF0EgEAAAAAEXQSAQAAAAARdBIBAAAAABF0EgEAgVRVVaW2bdtq8ODBkd8t\nXLhQrVq10kcffRT5neM4GjNmTIP7FhQUaP78+ZKkcDisfv36KScnR/3799ePfvQjffbZZ5F127Vr\np1WrVikvL095eXnq0KGDevTooby8PF100UWN6vrvf/+r/v37a9WqVZHf/epXv9LNN9+sdevWqW3b\ntpHHysvL02OPPRZZr7KyUq1atdJLL73U4DGPOuoo5eXladCgQRo3bpz27NkjSfrkk0+Um5ur9u3b\nx/NPCABAk+gkAgACq1evXnrnnXciyyUlJfrOd76jkpKSqPcLhUIKhUKRnx9//HGtXLlS7777rr72\nta/pu9/9boN1s7OzVVFRoYqKCo0dO1azZs1SRUWFXn755UaPfcwxx6ioqEg//OEPJUkbN27U3Llz\nNXPmTLmuq169ekUeq6KiQlddddVX1n/ssceqoqJC7777ro4//njNnTtXktSjRw9VVlbG+K8GAEB0\ndBIBAFbYs2ePli1bpjlz5ujJJ5+M6b6u60qS2rRpo1/+8pdav3693nvvva9cvzmjRo3SqaeequLi\nYk2ZMkWFhYU64YQTvrKGBQsW6P/9v/+n1157Tfv27WtyveHDh2vNmjVfkQgAgPjRSQQAWGHRokW6\n+OKLddppp+nkk09ucITxq9QdVZSkVq1aKScnRx9++GFC9RQVFelnP/uZtm3bpiuvvDLy+zVr1jQY\nbvrGG29Ikv7xj3+oZ8+e6tKli8LhsJ5//vlGj1lTU6OXX35Z2dnZCdUGAEA0rf0uAAAAL5SUlGjK\nlCmSpMsuu0wlJSUaPHhwgw5gfc39XvrqI4Utceqpp2rEiBGN5kP27NlTFRUVjdYvKSnRZZddJqm2\n/j//+c8aN26cJOmLL75QXl6eNm7cqKysLN18880J1wcAQHPoJAIAAm/79u0qKyvTqlWrFAqFVFNT\no1AopF/96lfq0KGDduzY0Wj9jh07NvlYNTU1eu+999S/f/+E62rVqlXUzmj955w/f76eeeYZ3Xvv\nvXJdV9u3b9fevXt13HHHqW3btqqoqNAXX3yhUaNGadGiRfre976XcH0AADSF4aYAgMB7+umndc01\n16iqqkpr167V+vXr1b17dy1dulR9+vTRpk2bIsNH161bp5UrVyo3Nzdy/7ojhwcOHNC0adN02mmn\npXRI5+LFi5Wbm6v169dr7dq1qqqq0rhx47RgwYIG67Vt21YPPvigfvazn3lytBMAgKbQSQQABN4T\nTzzR6Mja+PHj9cQTT+joo4/WY489ph/84AfKy8vTZZddpocffrjBZSOuvPJK5eTkaODAgfriiy+0\naNGiyN+aOhLYkqODza175JzE3/72t1HrP/IxcnNz1atXLz311FMtrgEAgFiEXL6KBAAEUFVVlcaM\nGRP1LKTpon379tq9e7ffZQAALMGRRABAILVu3VqfffaZBg8e7Hcpvvnkk0+Um5urzp07+10KAMAi\nHEkEACBO7733nq655poGvzvmmGP05ptv+lQRAACJo5MIAAAAAIhguCkAAAAAIIJOIgAAAAAggk4i\nAAAAACCCTiIAAAAAIIJOIgAAAAAg4v8D7Npmqbf+BecAAAAASUVORK5CYII=\n", "text": [ "" ] }, { "metadata": {}, "output_type": "display_data", "text": [ "" ] }, { "metadata": {}, "output_type": "display_data", "png": "iVBORw0KGgoAAAANSUhEUgAAA4gAAAFXCAYAAADgR2AdAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAASUVORK5CYII=\n", "text": [ "" ] }, { "metadata": {}, "output_type": "display_data", "text": [ "" ] }, { "metadata": {}, "output_type": "display_data", "png": "iVBORw0KGgoAAAANSUhEUgAAA4kAAAFXCAYAAAAPhQsjAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n", "text": [ "" ] } ], "prompt_number": 37 }, { "cell_type": "markdown", "metadata": {}, "source": [ "It is evident that many counties beat the 5.5% growth limit with some frequency." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Now it would be useful to understand the nature of the composite limit. The local property tax growth limitation imposed by TABOR is inflation plus new construction. We can use housing units growth to measure new construction and simply add in the CPI growth to create this dynamic limit. The Statewide Limit can be included trivially." ] }, { "cell_type": "code", "collapsed": false, "input": [ "#Create TABOR Limit variable\n", "co93['tabor_lim']=Series(co93['cpi_growth']+co93['hu_growth']).where(Series(co93['cpi_growth']+co93['hu_growth'])>0,0)\n", "\n", "#Create Statewide Limit variable\n", "co93['state_lim']=.055\n", "\n", "co93[['hu_num','cpi','cpi_growth','hu_growth','tabor_lim','state_lim']].head(15)" ], "language": "python", "metadata": {}, "outputs": [ { "html": [ "

		hu_num	cpi	cpi_growth	hu_growth	tabor_lim	state_lim
FNAME	AUDIT_YEAR
Adams	1993	109367	144.5	0.029936	0.011964	0.041900	0.055
1994	111800	148.2	0.025606	0.022246	0.047852	0.055
1995	114742	152.4	0.028340	0.026315	0.054655	0.055
1996	117691	156.9	0.029528	0.025701	0.055229	0.055
1997	120682	160.5	0.022945	0.025414	0.048359	0.055
1998	124925	163.0	0.015576	0.035159	0.050735	0.055
1999	129403	166.6	0.022086	0.035846	0.057931	0.055
2000	128717	172.2	0.033613	-0.005301	0.028312	0.055
2001	133917	177.1	0.028455	0.040399	0.068854	0.055
2002	139734	179.9	0.015810	0.043437	0.059248	0.055
2003	144635	184.0	0.022790	0.035074	0.057864	0.055
2004	149648	188.9	0.026630	0.034660	0.061290	0.055
2005	154113	195.3	0.033880	0.029837	0.063717	0.055
2006	157948	201.6	0.032258	0.024884	0.057142	0.055
2007	160651	207.3	0.028274	0.017113	0.045387	0.055

\n", "

15 rows \u00d7 6 columns

\n", "

" ], "metadata": {}, "output_type": "pyout", "prompt_number": 38, "text": [ " hu_num cpi cpi_growth hu_growth tabor_lim state_lim\n", "FNAME AUDIT_YEAR \n", "Adams 1993 109367 144.5 0.029936 0.011964 0.041900 0.055\n", " 1994 111800 148.2 0.025606 0.022246 0.047852 0.055\n", " 1995 114742 152.4 0.028340 0.026315 0.054655 0.055\n", " 1996 117691 156.9 0.029528 0.025701 0.055229 0.055\n", " 1997 120682 160.5 0.022945 0.025414 0.048359 0.055\n", " 1998 124925 163.0 0.015576 0.035159 0.050735 0.055\n", " 1999 129403 166.6 0.022086 0.035846 0.057931 0.055\n", " 2000 128717 172.2 0.033613 -0.005301 0.028312 0.055\n", " 2001 133917 177.1 0.028455 0.040399 0.068854 0.055\n", " 2002 139734 179.9 0.015810 0.043437 0.059248 0.055\n", " 2003 144635 184.0 0.022790 0.035074 0.057864 0.055\n", " 2004 149648 188.9 0.026630 0.034660 0.061290 0.055\n", " 2005 154113 195.3 0.033880 0.029837 0.063717 0.055\n", " 2006 157948 201.6 0.032258 0.024884 0.057142 0.055\n", " 2007 160651 207.3 0.028274 0.017113 0.045387 0.055\n", "\n", "[15 rows x 6 columns]" ] } ], "prompt_number": 38 }, { "cell_type": "markdown", "metadata": {}, "source": [ "Again looking at Adams County as the spot check, which limit is binding more often?" ] }, { "cell_type": "code", "collapsed": false, "input": [ "co93.ix['Adams'][['tabor_lim','state_lim']].plot(kind='line',title='TABOR vs Statewide Limit in Adams County')" ], "language": "python", "metadata": {}, "outputs": [ { "metadata": {}, "output_type": "pyout", "prompt_number": 39, "text": [ "" ] }, { "metadata": {}, "output_type": "display_data", "png": "iVBORw0KGgoAAAANSUhEUgAAA3gAAAFSCAYAAACgxn03AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAAQICgwQMAAACAAEGDBwAAAAAB4v8Auuj77CgM+ZkAAAAASUVORK5CYII=\n", "text": [ "" ] } ], "prompt_number": 39 }, { "cell_type": "markdown", "metadata": {}, "source": [ "It appears neither limit is likely to be dominant, but we can get a better global view by calculating the difference and looking at boxplots over time." ] }, { "cell_type": "code", "collapsed": false, "input": [ "Series(co93['tabor_lim']-co93['state_lim']).reset_index().boxplot(column=0,by='AUDIT_YEAR')" ], "language": "python", "metadata": {}, "outputs": [ { "metadata": {}, "output_type": "pyout", "prompt_number": 40, "text": [ "" ] }, { "metadata": {}, "output_type": "display_data", "png": "iVBORw0KGgoAAAANSUhEUgAAA48AAAFXCAYAAAACm3tkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAElFTkSuQmCC\n", "text": [ "" ] } ], "prompt_number": 40 }, { "cell_type": "markdown", "metadata": {}, "source": [ "This is interesting. The median position of the difference between the two limits is consistently negative, indicating that the TABOR limit is generally lower (and thus more binding) than the Statewide limit. There is, however, sufficient variation to preclude any notion that suggests the Statewide limit is inconsequential. Note that in 2009, CPI actually went down (which is the reason for the dramatically lower values)." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "\n", "\n", "$$\n", "\\text{if TABOR & Statewide:}\\\\\n", "\\hspace{4ex} \\text{limit_diff}=\\begin{cases}\n", "\\text{asmt_growth - min(TABOR,Statewide)}& \\text{if asmt_growth > min(TABOR,Statewide)},\\\\\n", "0& \\text{otherwise}.\n", "\\end{cases}\n", "$$\n", "$$\n", "\\text{elif TABOR & !Statewide:}\\\\\n", "\\hspace{4ex} \\text{limit_diff}=\\begin{cases}\n", "\\text{asmt_growth - TABOR}& \\text{if asmt_growth > TABOR},\\\\\n", "0& \\text{otherwise}.\n", "\\end{cases}\n", "$$\n", "$$\n", "\\text{elif !TABOR & Statewide:}\\\\\n", "\\hspace{4ex} \\text{limit_diff}=\\begin{cases}\n", "\\text{asmt_growth - Statewide}& \\text{if asmt_growth > Statewide},\\\\\n", "0& \\text{otherwise}.\n", "\\end{cases}\n", "$$\n", "$$\n", "\\text{else:}\\\\\n", "\\hspace{4ex} \\text{limit_diff}=0\n", "$$" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "There are two components to the limit dynamic. Within year, revenue growth (which can be proxied by assessment growth in a single year) cannot exceed the composite limit. But, there is also this persistent effect, that is a function of the distance between potential and realized revenue. Assessment growth increases this revenue potential quite independently of revenue growth *once the composite limit has been breached*. If we were to measure the revenue potential (rev_pot) as the product of the tax rate and the assessment base, we could determine whether or not the revenue potential is greater than the composite limit distance from the prior year's realized revenue. If it is, then the limit_diff must then be calculated as the difference between the composite limit and percentage growth in revenue required to exhaust the assessment base in a given year. If it isn't, then the composite limit is not binding. Assessment growth is not the right comparison, because it only tells the flow component of the story. The above method must be modified to also take account of the stock.\n", "\n", "One problem is, if we were to index assessment potential to compare it realized revenue, we would need a start year. That start year should probably be when TABOR was implemented, so ultimately I will need more assessment data. \n", "\n", "**UPDATE:** The assessment data has been updated to incorporate assessment manual data back to 1993. This was a manual extract from (gasp!) PDF storage. The Notebook responsible for the processing of this data was *asmt_add* (in the data directory).\n", "\n", "I can initialize the series with the median rate in the start year. If I were to just start the clock at zero, all counties would be in the same boat from an accumulated effect standpoint. Employing the median rate to simulate revenue potential permits variation in starting conditions. Note that the [annual reports](http://www.colorado.gov/cs/Satellite/DOLA-Main/CBON/1251591547549) have average mill rates by county, but the reports available online only go back to 2004. \n", "\n", "With the update, we can go back to 1993, but we still fall victim to the staggered start because the statewide growth limitation was in place prior to 1993. We need the median rates from 1993, but this requires a copy of the 1993 annual report (which is not available online). I have requested hard copies of reports in 1993 and 1987 (to accomodate the full range of finance data), but it will not be ready in time for the conference. Consequently, we will assume uniform starting points at the inception of TABOR. This is less than ideal, but it can be asserted that there was a categorical change at the inception of TABOR. It masks previous deficits, but the bias is less than it might have been with a start year of 1998.\n", "\n", "To operationalize this information, 1993 serves as $YEAR_0$. The 1993 assessment levels are treated as indicative of allowable revenues, an assumption we hope to remedy once the relevant data become available. In $YEAR_1$, we must calculate the distance between the growth in the assessment base and the baseline imposed by the composite limit. At this point, assessment value serves as our proxy for revenue yield. If the assessment value in $YEAR_1$ exceeds the sum of $YEAR_0$ assessment *and the growth permitted by the composite limit*, the TEL is actively constraining the county in question at a magnitude defined by said distance. If the assessment value growth does not exceed that which is allowable, then there is no active constraint imposed by the TEL.\n", "\n", "So far we have addressed only the flow portion of the equation. The stock portion becomes material in $YEAR_2$. At this point we again calculate the annual distance between the assessment growth and the growth allowed by the composite limit. However, to capture the stock effect, we employ the cumulative sum of these annual distances in an additional variable. This cumulative sum measures the extent to which fiscal capacity is hindered by the growth limitations imposed by TABOR and the Statewide Property Tax Revenue Limit. \n", "\n", "The implicit assumption that drives this model is that prior to TABOR, each countied levied taxes at rates that reflected the preferences of the county in question. Once TABOR was enacted,forced reductions in levies are actually wedges imposed by state law rather than county-specific preference. \n", "\n", "De-Brucing enters as a pressure release, eliminating one or both growth constraints. In the previous analysis (*CO_TEL_convergence*), TEL intensity was measured as a score, so the impact of De-Brucing was captured as a positive integer subtracted from the number of overlapping policies. In this analysis, De-Brucing will determine whether or not a given limit is present (TABOR or Statewide Limit on Property Tax Revenue) in the calculation of the composite limit. From a programmatic standpoint, the De-Brucing information must be entered upstream of the cumulative impact. We will leverage the same De-Brucing dictionary used in *CO_TEL_convergence*." ] }, { "cell_type": "code", "collapsed": false, "input": [ "debruce_dict={'Adams':[2002,1], 'Alamosa':[1997,1], 'Archuleta':[1994,0], 'Baca':[1995,0],\n", " 'Bent':[1993,0], 'Boulder':[2005,0], 'Chaffee':[1993,0], 'Cheyenne':[1996,0],\n", " 'Clear Creek':[1999,1], 'Conejos':[1996,0], 'Costilla':[1997,0], 'Crowley':[1994,0],\n", " 'Custer':[1997,1], 'Dolores':[2000,1], 'Douglas':[1997,1], 'Eagle':[1995,0],\n", " 'Garfield':[1994,0], 'Grand':[1996,1], 'Gunnison':[2000,0], 'Hinsdale':[2006,1],\n", " 'Huerfano':[2007,1], 'Jackson':[1999,1], 'Kiowa':[1997,0], 'Kit Carson':[1997,0],\n", " 'La Plata':[2002,1], 'Lake':[2011,1], 'Larimer':[1999,1], 'Lincoln':[1995,0],\n", " 'Logan':[1997,0], 'Mineral':[1995,0], 'Moffat':[1996,0], 'Montezuma':[2002,1],\n", " 'Morgan':[1996,0], 'Otero':[1995,0], 'Ouray':[2002,1], 'Phillips':[1995,0],\n", " 'Pitkin':[1994,0], 'Prowers':[1994,0], 'Rio Blanco':[1996,1], 'Rio Grande':[1999,1],\n", " 'Saguache':[1996,1], 'San Juan':[1995,1], 'San Miguel':[2005,0], 'Sedgwick':[1996,1],\n", " 'Summit':[1998,0], 'Teller':[1997,1], 'Washington':[1996,1], 'Yuma':[2004,1]}" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 41 }, { "cell_type": "markdown", "metadata": {}, "source": [ "We need to create indicator variables signifying whether or not the TABOR and SLPTR are active in a given county/year. The above dictionary provides the means to determine this. The first position in each value array indicates the year in which TABOR was waived. The second indicates whether or not the SLPTR was also waived along with TABOR. We can use this dictionary to generate both binary variables, with 1 indicating active and 100,000 indicating inactive. Why use 100,000 to indicate that the limit is inactive? Recall that our ultimate measure of interest is the cumulative distance between assessment value growth and the limited growth allowed by TELs, *when the distance is positive*. If allowable growth is so high that actual assessment growth has no hope of matching it, the limit is effectively inactive. (The inactive value was initially 100, but some limit values were too small for this to be a sufficient markup.)\n", "\n", "Note that all counties start with both limits active in 1993. The waiver years for each county turn the relevant limits off. For more discussion on the limitations of these data, see *CO_TEL_convergence*.\n", "\n", "The selection of the appropriate data is a little tricky. We want a slice of years *within* each county. The standard selection operators do not appear to capture such views easily, so rather than assign values to entire vector subsections at a time, we will assign values to individual year/county rows." ] }, { "cell_type": "code", "collapsed": false, "input": [ "#Create new active status variables\n", "co93['TABOR_active']=1\n", "co93['SL_active']=1\n", "\n", "#For each county...\n", "for key in debruce_dict.keys():\n", " #...if both limits were waived...\n", " if debruce_dict[key][1]==1:\n", " #...for each year from the waiver year until 2009\n", " for i in range(debruce_dict[key][0],2010):\n", " #...turn off both limits\n", " co93.loc[(key,i),'TABOR_active']=100000\n", " co93.loc[(key,i),'SL_active']=100000\n", " #...otherwise...\n", " else:\n", " #...for each year after the waiver year...\n", " for i in range(debruce_dict[key][0],2010):\n", " #...change only the TABOR status to off\n", " co93.loc[(key,i),'TABOR_active']=100000\n", " " ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 42 }, { "cell_type": "code", "collapsed": false, "input": [ "HTML(co93[['hu_num','cpi','cpi_growth','hu_growth','tabor_lim','state_lim','TABOR_active','SL_active']].head(15).to_html())" ], "language": "python", "metadata": {}, "outputs": [ { "html": [ "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "

		hu_num	cpi	cpi_growth	hu_growth	tabor_lim	state_lim	TABOR_active	SL_active
FNAME	AUDIT_YEAR
Adams	1993	109367	144.5	0.029936	0.011964	0.041900	0.055	1	1
1994	111800	148.2	0.025606	0.022246	0.047852	0.055	1	1
1995	114742	152.4	0.028340	0.026315	0.054655	0.055	1	1
1996	117691	156.9	0.029528	0.025701	0.055229	0.055	1	1
1997	120682	160.5	0.022945	0.025414	0.048359	0.055	1	1
1998	124925	163.0	0.015576	0.035159	0.050735	0.055	1	1
1999	129403	166.6	0.022086	0.035846	0.057931	0.055	1	1
2000	128717	172.2	0.033613	-0.005301	0.028312	0.055	1	1
2001	133917	177.1	0.028455	0.040399	0.068854	0.055	1	1
2002	139734	179.9	0.015810	0.043437	0.059248	0.055	100000	100000
2003	144635	184.0	0.022790	0.035074	0.057864	0.055	100000	100000
2004	149648	188.9	0.026630	0.034660	0.061290	0.055	100000	100000
2005	154113	195.3	0.033880	0.029837	0.063717	0.055	100000	100000
2006	157948	201.6	0.032258	0.024884	0.057142	0.055	100000	100000
2007	160651	207.3	0.028274	0.017113	0.045387	0.055	100000	100000

" ], "metadata": {}, "output_type": "pyout", "prompt_number": 43, "text": [ "" ] } ], "prompt_number": 43 }, { "cell_type": "markdown", "metadata": {}, "source": [ "Now we can define to new variables (*T_constraint* & *S_constraint*) that capture the interaction between active status and the actual value of the limit." ] }, { "cell_type": "code", "collapsed": false, "input": [ "co93['T_constraint']=co93['tabor_lim']*co93['TABOR_active']\n", "co93['S_constraint']=co93['state_lim']*co93['SL_active']\n", "\n", "co93[['tabor_lim','state_lim','TABOR_active','SL_active','S_constraint','T_constraint','d_total']].head(15)" ], "language": "python", "metadata": {}, "outputs": [ { "html": [ "

		tabor_lim	state_lim	TABOR_active	SL_active	S_constraint	T_constraint	d_total
FNAME	AUDIT_YEAR
Adams	1993	0.041900	0.055	1	1	0.055	0.041900	NaN
1994	0.047852	0.055	1	1	0.055	0.047852	0.031720
1995	0.054655	0.055	1	1	0.055	0.054655	0.077301
1996	0.055229	0.055	1	1	0.055	0.055229	0.046162
1997	0.048359	0.055	1	1	0.055	0.048359	0.103097
1998	0.050735	0.055	1	1	0.055	0.050735	0.052680
1999	0.057931	0.055	1	1	0.055	0.057931	0.158116
2000	0.028312	0.055	1	1	0.055	0.028312	0.060159
2001	0.068854	0.055	1	1	0.055	0.068854	0.219686
2002	0.059248	0.055	100000	100000	5500.000	5924.762967	0.012722
2003	0.057864	0.055	100000	100000	5500.000	5786.422218	0.079771
2004	0.061290	0.055	100000	100000	5500.000	6129.009531	0.043210
2005	0.063717	0.055	100000	100000	5500.000	6371.704340	0.093615
2006	0.057142	0.055	100000	100000	5500.000	5714.240263	0.031249
2007	0.045387	0.055	100000	100000	5500.000	4538.703666	0.065237

\n", "

15 rows \u00d7 7 columns

\n", "

" ], "metadata": {}, "output_type": "pyout", "prompt_number": 44, "text": [ " tabor_lim state_lim TABOR_active SL_active S_constraint T_constraint d_total\n", "FNAME AUDIT_YEAR \n", "Adams 1993 0.041900 0.055 1 1 0.055 0.041900 NaN\n", " 1994 0.047852 0.055 1 1 0.055 0.047852 0.031720\n", " 1995 0.054655 0.055 1 1 0.055 0.054655 0.077301\n", " 1996 0.055229 0.055 1 1 0.055 0.055229 0.046162\n", " 1997 0.048359 0.055 1 1 0.055 0.048359 0.103097\n", " 1998 0.050735 0.055 1 1 0.055 0.050735 0.052680\n", " 1999 0.057931 0.055 1 1 0.055 0.057931 0.158116\n", " 2000 0.028312 0.055 1 1 0.055 0.028312 0.060159\n", " 2001 0.068854 0.055 1 1 0.055 0.068854 0.219686\n", " 2002 0.059248 0.055 100000 100000 5500.000 5924.762967 0.012722\n", " 2003 0.057864 0.055 100000 100000 5500.000 5786.422218 0.079771\n", " 2004 0.061290 0.055 100000 100000 5500.000 6129.009531 0.043210\n", " 2005 0.063717 0.055 100000 100000 5500.000 6371.704340 0.093615\n", " 2006 0.057142 0.055 100000 100000 5500.000 5714.240263 0.031249\n", " 2007 0.045387 0.055 100000 100000 5500.000 4538.703666 0.065237\n", "\n", "[15 rows x 7 columns]" ] } ], "prompt_number": 44 }, { "cell_type": "markdown", "metadata": {}, "source": [ "We can now calculate the following values:\n", "\n", "1. The composite constraint (*comp_constraint*), which is just min(*S_constraint,T_constraint*)\n", "2. The annual distance (*intensity_flow*) between total assessment growth (*d_total*) and the composite constraint (*comp_constraint*)\n", "3. The cumulative distance (*intensity_stock*) between total assessment growth (*d_total*) and the composite constraint (*comp_constraint*)\n", "\n", "Note that each of these is a conditional assignment. *comp_constraint* is the minimum of two pre-existing Series. *intensity_flow* and *intensity_stock* only have non-zero values only if said values are positive. The latter is a bit tricky because it must reset whenever the potential is non-existent, ignoring any subsequent negative values. To spell this out more explicitly, there are three behavioral components for the stock values:\n", "\n", "1. If stock is already zero, negative flow values are irrelevant.\n", "2. If stock is positive, negative flow values must reduce said stock.\n", "3. Positive flow values always increase stock." ] }, { "cell_type": "code", "collapsed": false, "input": [ "print co93['pop_growth']" ], "language": "python", "metadata": {}, "outputs": [ { "output_type": "stream", "stream": "stdout", "text": [ "FNAME AUDIT_YEAR\n", "Adams 1993 0.014064\n", " 1994 0.027418\n", " 1995 0.031000\n", " 1996 0.024262\n", " 1997 0.030077\n", " 1998 0.026010\n", " 1999 0.024978\n", " 2000 0.101212\n", " 2001 -0.007132\n", " 2002 0.039080\n", " 2003 0.026325\n", " 2004 0.014359\n", " 2005 0.028956\n", " 2006 0.024332\n", " 2007 0.022149\n", "...\n", "Yuma 1995 0.018692\n", " 1996 0.005334\n", " 1997 0.021753\n", " 1998 0.014332\n", " 1999 0.002048\n", " 2000 0.005518\n", " 2001 0.005995\n", " 2002 0.001111\n", " 2003 0.010796\n", " 2004 -0.005790\n", " 2005 0.001807\n", " 2006 0.003107\n", " 2007 -0.003597\n", " 2008 -0.007921\n", " 2009 0.014352\n", "Name: pop_growth, Length: 1088, dtype: float64\n" ] } ], "prompt_number": 52 }, { "cell_type": "code", "collapsed": false, "input": [ "#Calculate composite constraint\n", "co93['comp_constraint']=co93['S_constraint'].where(co93['S_constraint'] < co93['T_constraint'],co93['T_constraint'])\n", "\n", "#Calculate annual distance\n", "co93['intensity_flow']=Series(co93['d_total']-co93['comp_constraint']).where(co93['d_total']>co93['comp_constraint'],0)\n", "\n", "#Initialize cumulative distance\n", "co93['intensity_stock']=co93['intensity_flow']\n", "\n", "##Calculate cumulative distance (split, apply, combine)\n", "#Create simple intensity difference\n", "co93['intensity_diff']=Series(co93['d_total']-co93['comp_constraint']).replace(NaN,0)\n", "\n", "#Create container for processed DF subsets\n", "cs_df_list=[]\n", "\n", "#Create list of counties\n", "cty_list=sorted(list(set(co93.index.get_level_values(level='FNAME'))))\n", "\n", "#For each county...\n", "for cty in cty_list:\n", " #...generate county-specific subset...\n", " cty_df=co93.ix[cty]\n", " #...for each year within county...\n", " for i in range(1993,2010):\n", " if (i==1993):\n", " cty_df.loc[i,'intensity_stock']=0\n", " else:\n", " if (cty_df.ix[i-1]['intensity_stock']==0) & (cty_df.ix[i]['intensity_diff']<0):\n", " cty_df.loc[i,'intensity_stock']=0\n", " elif (cty_df.ix[i-1]['intensity_stock']>0) & (cty_df.ix[i]['intensity_diff']<0): #Check this component\n", " cty_df.loc[i,'intensity_stock']=np.where(cty_df.ix[i-1]['intensity_stock']+cty_df.ix[i]['intensity_diff']>0,\\\n", " cty_df.ix[i-1]['intensity_stock']+cty_df.ix[i]['intensity_diff'],0)\n", " else:\n", " cty_df.loc[i,'intensity_stock']=cty_df.ix[i-1]['intensity_stock'] + cty_df.ix[i]['intensity_diff']\n", " #...re-integrate county name...\n", " cty_df['FNAME']=cty\n", " #...and throw processed DF subset into list\n", " cs_df_list.append(cty_df.reset_index())\n", "\n", "#Concatenate sets back together\n", "co_tel=pd.concat(cs_df_list).set_index(['FNAME','AUDIT_YEAR'])\n", "\n" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 46 }, { "cell_type": "code", "collapsed": false, "input": [ "HTML(co_tel.ix['Broomfield'][['tabor_lim','state_lim','TABOR_active','SL_active','S_constraint','T_constraint','d_total',\\\n", " 'comp_constraint','intensity_flow','intensity_diff','intensity_stock']].head(15).to_html())" ], "language": "python", "metadata": {}, "outputs": [ { "html": [ "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "

	tabor_lim	state_lim	TABOR_active	SL_active	S_constraint	T_constraint	d_total	comp_constraint	intensity_flow	intensity_diff	intensity_stock
AUDIT_YEAR
1993	0.000000	0.055	1	1	0.055	0.000000	NaN	0.000	0.000000	0.000000	0.000000
1994	0.000000	0.055	1	1	0.055	0.000000	NaN	0.000	0.000000	0.000000	0.000000
1995	0.000000	0.055	1	1	0.055	0.000000	NaN	0.000	0.000000	0.000000	0.000000
1996	0.000000	0.055	1	1	0.055	0.000000	NaN	0.000	0.000000	0.000000	0.000000
1997	0.000000	0.055	1	1	0.055	0.000000	NaN	0.000	0.000000	0.000000	0.000000
1998	0.000000	0.055	1	1	0.055	0.000000	NaN	0.000	0.000000	0.000000	0.000000
1999	0.000000	0.055	1	1	0.055	0.000000	NaN	0.000	0.000000	0.000000	0.000000
2000	0.000000	0.055	1	1	0.055	0.000000	NaN	0.000	0.000000	0.000000	0.000000
2001	0.070393	0.055	1	1	0.055	0.070393	NaN	0.055	0.000000	0.000000	0.000000
2002	0.070606	0.055	1	1	0.055	0.070606	0.102509	0.055	0.047509	0.047509	0.047509
2003	0.080658	0.055	1	1	0.055	0.080658	0.092171	0.055	0.037171	0.037171	0.084680
2004	0.071154	0.055	1	1	0.055	0.071154	-0.003857	0.055	0.000000	-0.058857	0.025823
2005	0.103974	0.055	1	1	0.055	0.103974	0.001123	0.055	0.000000	-0.053877	0.000000
2006	0.082902	0.055	1	1	0.055	0.082902	0.028180	0.055	0.000000	-0.026820	0.000000
2007	0.078859	0.055	1	1	0.055	0.078859	0.115620	0.055	0.060620	0.060620	0.060620

" ], "metadata": {}, "output_type": "pyout", "prompt_number": 54, "text": [ "" ] } ], "prompt_number": 54 }, { "cell_type": "markdown", "metadata": {}, "source": [ "What does the dynamic of flow versus stock TEL constraints look like? Again, Adams County serves as our test bed." ] }, { "cell_type": "code", "collapsed": false, "input": [ "fig,axes=plt.subplots(2)\n", "co_tel.ix['Adams']['intensity_flow'].plot(kind='bar',ax=axes[0])\n", "co_tel.ix['Adams']['intensity_stock'].plot(kind='line',ax=axes[1])" ], "language": "python", "metadata": {}, "outputs": [ { "metadata": {}, "output_type": "pyout", "prompt_number": 55, "text": [ "" ] }, { "metadata": {}, "output_type": "display_data", "png": "iVBORw0KGgoAAAANSUhEUgAAA3gAAAFICAYAAAACi1zKAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nsMEjIiIiIiJyEWzwiIiIiIiIXMT/A2bo54jKQOmlAAAAAElFTkSuQmCC\n", "text": [ "" ] } ], "prompt_number": 55 }, { "cell_type": "markdown", "metadata": {}, "source": [ "The Gallagher Amendment is comparatively easy to model. Given that the restriction sets assessment rates based upon the statewide ratio of residential and non-residential property value, differential impacts by county can be modeled as the county-level ratios. To the extent that an individual county features a different mix of property, the impact of GA on said county will differ from the statewide effect (and other counties). Since assessment rates apply statewide (and thus are uniform modifiers of market value), we can model this ratio as residual over non-residual assessment value." ] }, { "cell_type": "code", "collapsed": false, "input": [ "#Generate property ratio variable\n", "co_tel['prop_ratio']=co_tel['resid']/co_tel['non_resid']\n", "\n", "#Observe ratio distribution over time\n", "co_tel['prop_ratio'].reset_index().boxplot(column='prop_ratio',by='AUDIT_YEAR')" ], "language": "python", "metadata": {}, "outputs": [ { "metadata": {}, "output_type": "pyout", "prompt_number": 56, "text": [ "" ] }, { "metadata": {}, "output_type": "display_data", "png": "iVBORw0KGgoAAAANSUhEUgAAA4EAAAFXCAYAAAAcUkvXAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n", "text": [ "" ] } ], "prompt_number": 56 }, { "cell_type": "markdown", "metadata": {}, "source": [ "***NOTE:*** Check Regional Directory at BEA for county level business investment. If I can't find it, ask Wendy about it in an email." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We are now in a position to restrict the DataFrame to the variables we need.\n", "\n", "+ Per Capita Annual Payroll (*pcap*)\n", "+ Per Capita Revenue (*pcrev*)\n", "+ LISAs (*w_rook,w_queen,w_db_b,w_db_b,w_kern* for all LISA-base suffixes) \n", "+ Gross State Product (*gsp*)\n", "+ State Unemployment Rate (*st_unempr*)\n", "+ Housing Permits/Housing Units (*permits/hu_num*)\n", "+ Vacancy Rate (*vac_rate*)\n", "+ Labor Force/Population (*labforce/CTY_POP*)\n", "+ Lagged Population Growth (lag(diff(*CTY_POP*)/(*CTY_POP*)))\n", "+ TEL Intensity Flow (*intensity_flow*)\n", "+ TEL Intensity Stock (*intensity_stock*)\n", "+ Gallagher Ratio (*prop_ratio*)\n", "+ Integovernmental transfers (*pcintgov*)" ] }, { "cell_type": "code", "collapsed": false, "input": [ "#Define desired columns\n", "tel_cols=['pcap','pcrev','w_db_b_g','w_db_b_g_ap','w_db_b_i','w_db_b_i_ap', \\\n", " 'w_db_c_g','w_db_c_g_ap','w_db_c_i','w_db_c_i_ap','w_kern_g','w_kern_g_ap','w_kern_i', \\\n", " 'w_kern_i_ap','w_queen_g','w_queen_g_ap','w_queen_i','w_queen_i_ap','w_rook_g','w_rook_g_ap',\\\n", " 'w_rook_i','w_rook_i_ap','gsp','st_unempr','permits','hu_num','vac_rate','labforce',\\\n", " 'CTY_POP','intensity_flow','intensity_stock','prop_ratio','pcintgov','pop_growth']\n", "\n", "#Subset to desired columns\n", "co_TEL=co_tel.reindex(columns=tel_cols)\n", "\n", "#Create new variables\n", "co_TEL['permit_rate']=co_TEL['permits']/co_TEL['hu_num']\n", "co_TEL['lfp_rat']=co_TEL['labforce']/co_TEL['CTY_POP']\n", "# co_TEL['pop_growth']=co_TEL['CTY_POP'].pct_change()\n", "co_TEL['lpop_growth']=co_TEL['pop_growth'].shift(1)\n", "\n", "#Write final set to disk\n", "co_TEL.to_csv(workdir+'co_TEL_df.csv')\n", "\n", "co_TEL.head().T" ], "language": "python", "metadata": {}, "outputs": [ { "html": [ "

FNAME	Adams
AUDIT_YEAR	1993	1994	1995	1996	1997
pcap	8363.248319	8420.640636	8659.319256	9692.246662	9903.654147
pcrev	613.581370	619.581576	625.547804	631.738963	597.750806
w_db_b_g	-0.666336	-0.702718	-0.685571	-0.690522	-0.636007
w_db_b_g_ap	-0.218198	-0.218817	-0.216704	-0.218042	-0.214792
w_db_b_i	0.726357	0.776141	0.750504	0.754434	0.732544
w_db_b_i_ap	0.097566	0.099612	0.102275	0.100128	0.102452
w_db_c_g	-0.863283	-0.922535	-0.925116	-0.917968	-0.895004
w_db_c_g_ap	-0.208308	-0.206978	-0.203922	-0.203195	-0.198834
w_db_c_i	0.912518	0.988179	0.992127	0.977454	1.010452
w_db_c_i_ap	0.093739	0.095023	0.093708	0.099272	0.092291
w_kern_g	-1.788854	-1.929129	-1.925412	-1.831087	-1.852637
w_kern_g_ap	-0.423035	-0.422102	-0.421102	-0.420230	-0.418237
w_kern_i	1.778028	1.981237	1.985268	1.864233	2.019240
w_kern_i_ap	0.155047	0.158922	0.158193	0.155329	0.154766
w_queen_g	-0.805457	-0.845043	-0.772144	-0.714661	-0.659345
w_queen_g_ap	-0.309126	-0.307735	-0.307370	-0.310988	-0.306477
w_queen_i	0.741493	0.842483	0.763211	0.690906	0.689241
w_queen_i_ap	0.109414	0.118862	0.111866	0.109368	0.110782
w_rook_g	-0.791534	-0.833280	-0.759176	-0.702119	-0.646718
w_rook_g_ap	-0.304737	-0.303316	-0.302573	-0.305390	-0.300630
w_rook_i	0.730110	0.818662	0.748184	0.680452	0.675022
w_rook_i_ap	0.118120	0.110652	0.113251	0.105931	0.111892
gsp	103457.000000	110529.000000	116821.000000	123144.000000	157345.000000
st_unempr	5.300000	4.300000	4.000000	4.200000	3.400000
permits	2439.000000	2906.000000	2944.000000	2764.000000	4155.000000
hu_num	109367.000000	111800.000000	114742.000000	117691.000000	120682.000000
vac_rate	6.000000	5.900000	5.500000	5.500000	5.100000
labforce	151092.000000	158830.000000	165754.000000	169548.000000	173611.000000
CTY_POP	281130.000000	288838.000000	297792.000000	305017.000000	314191.000000
intensity_flow	0.000000	0.000000	0.022646	0.000000	0.054739
intensity_stock	0.000000	0.000000	0.022646	0.013808	0.068546
prop_ratio	0.760530	0.788998	0.812398	0.816137	0.838724
pcintgov	282.333128	265.877196	251.412599	243.214670	210.599472
pop_growth	0.014064	0.027418	0.031000	0.024262	0.030077
permit_rate	0.022301	0.025993	0.025658	0.023485	0.034429
lfp_rat	0.537445	0.549893	0.556610	0.555864	0.552565
lpop_growth	NaN	0.014064	0.027418	0.031000	0.024262

\n", "

37 rows \u00d7 5 columns

\n", "

" ], "metadata": {}, "output_type": "pyout", "prompt_number": 57, "text": [ "FNAME Adams \n", "AUDIT_YEAR 1993 1994 1995 1996 1997\n", "pcap 8363.248319 8420.640636 8659.319256 9692.246662 9903.654147\n", "pcrev 613.581370 619.581576 625.547804 631.738963 597.750806\n", "w_db_b_g -0.666336 -0.702718 -0.685571 -0.690522 -0.636007\n", "w_db_b_g_ap -0.218198 -0.218817 -0.216704 -0.218042 -0.214792\n", "w_db_b_i 0.726357 0.776141 0.750504 0.754434 0.732544\n", "w_db_b_i_ap 0.097566 0.099612 0.102275 0.100128 0.102452\n", "w_db_c_g -0.863283 -0.922535 -0.925116 -0.917968 -0.895004\n", "w_db_c_g_ap -0.208308 -0.206978 -0.203922 -0.203195 -0.198834\n", "w_db_c_i 0.912518 0.988179 0.992127 0.977454 1.010452\n", "w_db_c_i_ap 0.093739 0.095023 0.093708 0.099272 0.092291\n", "w_kern_g -1.788854 -1.929129 -1.925412 -1.831087 -1.852637\n", "w_kern_g_ap -0.423035 -0.422102 -0.421102 -0.420230 -0.418237\n", "w_kern_i 1.778028 1.981237 1.985268 1.864233 2.019240\n", "w_kern_i_ap 0.155047 0.158922 0.158193 0.155329 0.154766\n", "w_queen_g -0.805457 -0.845043 -0.772144 -0.714661 -0.659345\n", "w_queen_g_ap -0.309126 -0.307735 -0.307370 -0.310988 -0.306477\n", "w_queen_i 0.741493 0.842483 0.763211 0.690906 0.689241\n", "w_queen_i_ap 0.109414 0.118862 0.111866 0.109368 0.110782\n", "w_rook_g -0.791534 -0.833280 -0.759176 -0.702119 -0.646718\n", "w_rook_g_ap -0.304737 -0.303316 -0.302573 -0.305390 -0.300630\n", "w_rook_i 0.730110 0.818662 0.748184 0.680452 0.675022\n", "w_rook_i_ap 0.118120 0.110652 0.113251 0.105931 0.111892\n", "gsp 103457.000000 110529.000000 116821.000000 123144.000000 157345.000000\n", "st_unempr 5.300000 4.300000 4.000000 4.200000 3.400000\n", "permits 2439.000000 2906.000000 2944.000000 2764.000000 4155.000000\n", "hu_num 109367.000000 111800.000000 114742.000000 117691.000000 120682.000000\n", "vac_rate 6.000000 5.900000 5.500000 5.500000 5.100000\n", "labforce 151092.000000 158830.000000 165754.000000 169548.000000 173611.000000\n", "CTY_POP 281130.000000 288838.000000 297792.000000 305017.000000 314191.000000\n", "intensity_flow 0.000000 0.000000 0.022646 0.000000 0.054739\n", "intensity_stock 0.000000 0.000000 0.022646 0.013808 0.068546\n", "prop_ratio 0.760530 0.788998 0.812398 0.816137 0.838724\n", "pcintgov 282.333128 265.877196 251.412599 243.214670 210.599472\n", "pop_growth 0.014064 0.027418 0.031000 0.024262 0.030077\n", "permit_rate 0.022301 0.025993 0.025658 0.023485 0.034429\n", "lfp_rat 0.537445 0.549893 0.556610 0.555864 0.552565\n", "lpop_growth NaN 0.014064 0.027418 0.031000 0.024262\n", "\n", "[37 rows x 5 columns]" ] } ], "prompt_number": 57 }, { "cell_type": "heading", "level": 2, "metadata": {}, "source": [ "Estimating the Impact of Tax and Expenditure Limitation Constraints" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "This section will implement the actual models of concern. Following the example set by *CO_TEL_convergence*, we will start by implementing the same models with different specifications. In this case, we will be using the models described at the beginning of the **Model Specification** section above.\n", "\n", "+ **Ordinary Least Squares** - OLS will serve as a non-spatial pooled estimate\n", "\n", " $$y=X\\beta+\\epsilon$$\n", "\n", "\n", "+ **Spatial Error Model** - The error term in this model seeks to mitigate spatial dependency across counties\n", "\n", " $y=X\\beta+\\epsilon$ and \n", " \n", " $\\epsilon=\\gamma W \\epsilon + u$ \n", " \n", " becomes $y=\\gamma Wy + X\\beta-\\gamma WX\\beta + \\epsilon$\n", "\n", "\n", "+ **Spatial Lag Model** - This model explicitly incorporates the weighted average of neighborhood values\n", "\n", " $y = \\rho Wy + X\\beta + \\epsilon$\n", " \n", " becomes $y=(I - \\rho W)^{-1} X\\beta + (I - \\rho W)^{-1} \\epsilon$\n", " \n", "\n", "+ **Fixed Effect Spatial Lag Model** - This model both incorporates the weighted average of neighborhood values and the value of the observation in the preceding time period\n", "\n", " $$y=\\lambda (I_T \\otimes W_N)y + (\\iota_T \\otimes I_N)\\mu + X\\beta + \\epsilon$$ " ] }, { "cell_type": "heading", "level": 3, "metadata": {}, "source": [ "Ordinary Least Squares" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The first model to be run ignores spatial association and the panel structure of the data. It is meant to serve purely as a baseline to better understand the impact of incorporating these features into the estimation process. \n", "\n", "A few items of note regarding modeling with [PySAL](http://pythonhosted.org/PySAL/index.html):\n", "\n", "+ The dependent variable, $y$, must be structured as an $n \\times 1$ numpy array\n", "+ The regressor matrix, $X$, must be structured as a $n \\times j$ numpy array, where $j$ is the number of independent variables\n", "+ As a consequence of working with numpy arrays, labeling information must be explicitly entered as parameters to the model\n", "\n", "In keeping with the above practice, all weight matrices will be tested. " ] }, { "cell_type": "code", "collapsed": false, "input": [ "#Define model dependents\n", "y_econ='pcap'\n", "y_rev='pcrev'\n", "y_clust_econ=['w_rook_i_ap','w_queen_i_ap','w_db_b_i_ap','w_db_c_i_ap','w_kern_i_ap',\\\n", " 'w_rook_g_ap','w_queen_g_ap','w_db_b_g_ap','w_db_c_g_ap','w_kern_g_ap']\n", "y_clust_rev=['w_rook_i','w_queen_i','w_db_b_i','w_db_c_i','w_kern_i',\\\n", " 'w_rook_g','w_queen_g','w_db_b_g','w_db_c_g','w_kern_g']\n", "\n", "#Set list of regressors for each model\n", "X_econ=['gsp','st_unempr','permit_rate','vac_rate','intensity_stock','prop_ratio']\n", "X_rev=['gsp','lpop_growth','st_unempr','permit_rate','vac_rate','pcap','intensity_stock','prop_ratio']\n", "X_clust=['gsp','lpop_growth','st_unempr','permit_rate','vac_rate','pcap','intensity_stock','prop_ratio']\n", "\n", "#Define variable that defines non-null subset\n", "null_var='pcap'\n", "\n", "#Define function to check for null values in regressor matrix\n", "def count_missing(frame):\n", " return (frame.shape[0] * frame.shape[1]) - frame.count().sum()\n", "\n", "def ols_pool(df,df_name,yvar,x_list,null_check):\n", " \"\"\"This function is utility to facilitate looping through all weighting schemes.\n", " Note: data are subset to provide only non-null values in relative annual payroll per capita.\n", " The collection of annual payroll data in one county did not begin until 2002.\"\"\"\n", " #Set dependent variable\n", " #print '\\nSETTING DEPENDENT:',yvar\n", " y=df[yvar][df[null_check].notnull()].values\n", " \n", " #Adhere to the structure constraints imposed by PySAL\n", " y.shape=(len(df[yvar][df[null_check].notnull()]),1)\n", " \n", " #Create container for regressors\n", " X=[]\n", " \n", " #Add regressors to container in turn\n", " #print '\\nBUILDING REGRESSOR MATRIX'\n", " for reg in x_list:\n", " #print '\\t',reg\n", " X.append(df[reg][df[null_check].notnull()].values)\n", " \n", " #Check for null values\n", " #print DataFrame(X)\n", " #null_series=DataFrame(X).groupby(DataFrame(X).index).apply(count_missing)\n", " #print null_series\n", " #print null_series[null_series>0]\n", " \n", " #Convert from j x n to n x j (again to adhere to PySAL structure constraints\n", " #print '\\nTRANSPOSING REGRESSOR MATRIX'\n", " X=np.array(X).T\n", " #print X\n", " \n", " #Run model\n", " #print '\\nIMPLEMENTING OLS\\n****************\\n'\n", " ols=pysal.spreg.ols.OLS(y,X,name_y=yvar,name_x=x_list,name_ds=df_name)\n", " return ols\n", "print '***TEST RUN - Dependent = Annual Payroll Per Capita (Economic Capacity)***\\n'\n", "print ols_pool(co_TEL,'co_TEL',y_econ,X_econ,null_var).summary" ], "language": "python", "metadata": {}, "outputs": [ { "output_type": "stream", "stream": "stdout", "text": [ "***TEST RUN - Dependent = Annual Payroll Per Capita (Economic Capacity)***\n", "\n", "REGRESSION\n", "----------\n", "SUMMARY OF OUTPUT: ORDINARY LEAST SQUARES\n", "-----------------------------------------\n", "Data set : co_TEL\n", "Dependent Variable : pcap Number of Observations: 1079\n", "Mean dependent var : 8171.1330 Number of Variables : 7\n", "S.D. dependent var : 7001.7488 Degrees of Freedom : 1072\n", "R-squared : 0.1980\n", "Adjusted R-squared : 0.1935\n", "Sum squared residual:42385461650.565 F-statistic : 44.1042\n", "Sigma-square :39538676.913 Prob(F-statistic) : 2.528e-48\n", "S.E. of regression : 6287.979 Log likelihood : -10964.883\n", "Sigma-square ML :39282170.204 Akaike info criterion : 21943.767\n", "S.E of regression ML: 6267.5490 Schwarz criterion : 21978.653\n", "\n", "------------------------------------------------------------------------------------\n", " Variable Coefficient Std.Error t-Statistic Probability\n", "------------------------------------------------------------------------------------\n", " CONSTANT -1038.4380811 1077.4107829 -0.9638274 0.3353498\n", " gsp 0.0283579 0.0049955 5.6767529 0.0000000\n", " intensity_stock 2952.8366182 728.9709715 4.0506916 0.0000547\n", " permit_rate 27028.5993450 12261.0085725 2.2044352 0.0277056\n", " prop_ratio 4930.2353149 484.8311869 10.1689731 0.0000000\n", " st_unempr 93.3437244 154.7244312 0.6032901 0.5464433\n", " vac_rate -19.7450398 10.6551202 -1.8531034 0.0641422\n", "------------------------------------------------------------------------------------\n", "\n", "REGRESSION DIAGNOSTICS\n", "MULTICOLLINEARITY CONDITION NUMBER 15.918559\n", "\n", "TEST ON NORMALITY OF ERRORS\n", "TEST DF VALUE PROB\n", "Jarque-Bera 2 842.092908 0.0000000\n", "\n", "DIAGNOSTICS FOR HETEROSKEDASTICITY\n", "RANDOM COEFFICIENTS\n", "TEST DF VALUE PROB\n", "Breusch-Pagan test 6 229.438639 0.0000000\n", "Koenker-Bassett test 6 89.059755 0.0000000\n", "================================ END OF REPORT =====================================" ] }, { "output_type": "stream", "stream": "stdout", "text": [ "\n" ] } ], "prompt_number": 31 }, { "cell_type": "markdown", "metadata": {}, "source": [ "The model executes, so let's pull some information into DataFrames for all dependents. Note that lagged population growth does not have non-null values until 1995." ] }, { "cell_type": "code", "collapsed": false, "input": [ "co_TEL95=co_TEL[co_TEL.index.get_level_values(level='AUDIT_YEAR').isin(range(1995,2010))]" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 32 }, { "cell_type": "markdown", "metadata": {}, "source": [ "We need a general workflow for comparing estimates across models. The function below captures critical information in DataFrame." ] }, { "cell_type": "code", "collapsed": false, "input": [ "def ols_df(model,model_name,reg_list):\n", " print '***RUNNING',model_name,'***\\n'\n", " \n", " #Create containers for estimates and p-values\n", " est_list=[]\n", " pval_list=[]\n", " col_list=[]\n", " \n", " #Populate estimate and p-value containers\n", " for col in model.name_x:\n", " col_list.append(col)\n", " for est in model.betas.tolist():\n", " est_list.append(est[0])\n", " for tstat_tuple in model.t_stat:\n", " pval_list.append(tstat_tuple[1])\n", " \n", " #Generate model DF\n", " ols_df=DataFrame({'var':col_list[1:],\n", " 'est':est_list[1:],\n", " 'pval':pval_list[1:]})\n", " \n", " #Include R^2\n", " ols_df['r2']=model.r2\n", " \n", " #Include model name\n", " ols_df['model']=model_name\n", " \n", " return ols_df\n", "\n", "#Capture summary info\n", "ols_econ_summary=ols_df(ols_pool(co_TEL,'co_TEL',y_econ,X_econ,null_var),'ols_econ',X_econ)\n", "\n", "ols_econ_summary" ], "language": "python", "metadata": {}, "outputs": [ { "output_type": "stream", "stream": "stdout", "text": [ "***RUNNING ols_econ ***\n", "\n" ] }, { "html": [ "

	est	pval	var	r2	model
0	0.028358	1.766288e-08	gsp	0.19798	ols_econ
1	93.343724	5.464433e-01	st_unempr	0.19798	ols_econ
2	27028.599345	2.770561e-02	permit_rate	0.19798	ols_econ
3	-19.745040	6.414218e-02	vac_rate	0.19798	ols_econ
4	2952.836618	5.474689e-05	intensity_stock	0.19798	ols_econ
5	4930.235315	2.976822e-23	prop_ratio	0.19798	ols_econ

\n", "

" ], "metadata": {}, "output_type": "pyout", "prompt_number": 33, "text": [ " est pval var r2 model\n", "0 0.028358 1.766288e-08 gsp 0.19798 ols_econ\n", "1 93.343724 5.464433e-01 st_unempr 0.19798 ols_econ\n", "2 27028.599345 2.770561e-02 permit_rate 0.19798 ols_econ\n", "3 -19.745040 6.414218e-02 vac_rate 0.19798 ols_econ\n", "4 2952.836618 5.474689e-05 intensity_stock 0.19798 ols_econ\n", "5 4930.235315 2.976822e-23 prop_ratio 0.19798 ols_econ" ] } ], "prompt_number": 33 }, { "cell_type": "markdown", "metadata": {}, "source": [ "We can run the revenue model the same way." ] }, { "cell_type": "code", "collapsed": false, "input": [ "ols_rev_summary=ols_df(ols_pool(co_TEL95,'co_TEL',y_rev,X_rev,null_var),'ols_rev',X_rev)\n", "\n", "ols_rev_summary" ], "language": "python", "metadata": {}, "outputs": [ { "output_type": "stream", "stream": "stdout", "text": [ "***RUNNING ols_rev ***\n", "\n" ] }, { "html": [ "

	est	pval	var	r2	model
0	0.000999	9.998783e-02	gsp	0.599852	ols_rev
1	-559.687132	4.105274e-01	lpop_growth	0.599852	ols_rev
2	19.308982	1.785172e-01	st_unempr	0.599852	ols_rev
3	-4881.559301	2.155566e-04	permit_rate	0.599852	ols_rev
4	32.375174	1.032227e-154	vac_rate	0.599852	ols_rev
5	0.044216	1.253257e-52	pcap	0.599852	ols_rev
6	-385.677194	2.772749e-09	intensity_stock	0.599852	ols_rev
7	-622.994861	1.235285e-37	prop_ratio	0.599852	ols_rev

\n", "

" ], "metadata": {}, "output_type": "pyout", "prompt_number": 34, "text": [ " est pval var r2 model\n", "0 0.000999 9.998783e-02 gsp 0.599852 ols_rev\n", "1 -559.687132 4.105274e-01 lpop_growth 0.599852 ols_rev\n", "2 19.308982 1.785172e-01 st_unempr 0.599852 ols_rev\n", "3 -4881.559301 2.155566e-04 permit_rate 0.599852 ols_rev\n", "4 32.375174 1.032227e-154 vac_rate 0.599852 ols_rev\n", "5 0.044216 1.253257e-52 pcap 0.599852 ols_rev\n", "6 -385.677194 2.772749e-09 intensity_stock 0.599852 ols_rev\n", "7 -622.994861 1.235285e-37 prop_ratio 0.599852 ols_rev" ] } ], "prompt_number": 34 }, { "cell_type": "markdown", "metadata": {}, "source": [ "We can also roll through each of the LISAs as dependents in this fashion. Furthermore, we can capture each resultant DataFrame in a list." ] }, { "cell_type": "code", "collapsed": false, "input": [ "#Create container for summary DFs\n", "ols_summ_list=[ols_econ_summary,ols_rev_summary]\n", "\n", "##Capture all models for revenue clustering\n", "#For each clustering dependent variable...\n", "for i in range(len(y_clust_rev)):\n", " #Run pooled ols and capture summary info\n", " summary_df_rev=ols_df(ols_pool(co_TEL95,'co_TEL',y_clust_rev[i],X_clust,null_var),'ols_'+y_clust_rev[i],X_clust)\n", " summary_df_econ=ols_df(ols_pool(co_TEL95,'co_TEL',y_clust_econ[i],X_clust,null_var),'ols_'+y_clust_econ[i],X_clust)\n", " #Dump summary info in the DF list\n", " ols_summ_list.append(summary_df_rev)\n", " ols_summ_list.append(summary_df_econ)\n", "\n", "#Concatenate all sets together\n", "ols_results=pd.concat(ols_summ_list).set_index(['var','model']).sortlevel(0)\n", "\n", "HTML(ols_results.head(15).to_html())" ], "language": "python", "metadata": {}, "outputs": [ { "output_type": "stream", "stream": "stdout", "text": [ "***RUNNING ols_w_rook_i ***\n", "\n", "***RUNNING ols_w_rook_i_ap ***\n", "\n", "***RUNNING ols_w_queen_i ***\n", "\n", "***RUNNING ols_w_queen_i_ap ***\n", "\n", "***RUNNING ols_w_db_b_i ***\n", "\n", "***RUNNING ols_w_db_b_i_ap ***\n", "\n", "***RUNNING ols_w_db_c_i ***\n", "\n", "***RUNNING ols_w_db_c_i_ap ***\n", "\n", "***RUNNING ols_w_kern_i ***\n", "\n", "***RUNNING ols_w_kern_i_ap ***\n", "\n", "***RUNNING" ] }, { "output_type": "stream", "stream": "stdout", "text": [ " ols_w_rook_g ***\n", "\n", "***RUNNING ols_w_rook_g_ap ***\n", "\n", "***RUNNING ols_w_queen_g ***\n", "\n", "***RUNNING ols_w_queen_g_ap ***\n", "\n", "***RUNNING ols_w_db_b_g ***\n", "\n", "***RUNNING ols_w_db_b_g_ap ***\n", "\n", "***RUNNING ols_w_db_c_g ***\n", "\n", "***RUNNING ols_w_db_c_g_ap ***\n", "\n", "***RUNNING ols_w_kern_g ***\n", "\n", "***RUNNING" ] }, { "output_type": "stream", "stream": "stdout", "text": [ " ols_w_kern_g_ap ***\n", "\n" ] }, { "html": [ "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "

		est	pval	r2
var	model
gsp	ols_econ	0.028358	1.766288e-08	0.197980
ols_rev	0.000999	9.998783e-02	0.599852
ols_w_db_b_g	-0.000001	2.389941e-01	0.269747
ols_w_db_b_g_ap	0.000002	3.515882e-02	0.060021
ols_w_db_b_i	0.000001	3.468063e-01	0.161128
ols_w_db_b_i_ap	-0.000001	3.987868e-01	0.029094
ols_w_db_c_g	-0.000001	2.570447e-01	0.261938
ols_w_db_c_g_ap	0.000002	3.169784e-02	0.060634
ols_w_db_c_i	0.000001	3.826184e-01	0.179876
ols_w_db_c_i_ap	-0.000001	3.229244e-01	0.031176
ols_w_kern_g	-0.000001	6.087666e-01	0.321585
ols_w_kern_g_ap	0.000001	2.129177e-01	0.060268
ols_w_kern_i	0.000001	6.247623e-01	0.132709
ols_w_kern_i_ap	-0.000001	7.246783e-01	0.010119
ols_w_queen_g	-0.000002	1.919297e-01	0.199230

" ], "metadata": {}, "output_type": "pyout", "prompt_number": 35, "text": [ "" ] } ], "prompt_number": 35 }, { "cell_type": "markdown", "metadata": {}, "source": [ "We will wait to explore the results until after the other models have been run." ] }, { "cell_type": "heading", "level": 3, "metadata": {}, "source": [ "Spatial Lag Model" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Spatial Lag Models (SLMs) seek to model the neighborhood impacts directly. The time series equivalent of this task would be an autoregressive specification seeking to model the impact of preceding observations $y_{t-1},y_{t-2},...,y_{t-n}$ on the current observation $y_t$. \n", "\n", "$$Y_t=\\alpha + \\sum_{i=1}^p \\rho_i Y_{t-i} + X\\beta + \\epsilon_t$$\n", "\n", "In the spatial context, this \"autoregressive\" impact captures the weighted average of the values for all neighboring geographies, $y_j \\forall j \\neq i$, in the neighborhood (defined by $W$) of geography unit $y_i$.\n", "\n", "$$y =\\alpha + \\rho Wy + X\\beta + \\epsilon$$\n", "\n", "The SLM is a cross-sectional model that must be run for each year. First, however, we need to generate some weight matrices. We will use the same approach taken in *CO_TEL_convergence*. Note that the distance based thresholds, *distance band (db)* and *kernel (kern)*, required a minimum threshold distance. That threshold will be set to ensure all geographies have at least one neighbor." ] }, { "cell_type": "code", "collapsed": false, "input": [ "#Establish shapefile location\n", "shp_dir='/home/choct155/Google Drive/Dissertation/Data/spatial/US/'\n", "\n", "#Generate rook weights\n", "w_rook=pysal.rook_from_shapefile(shp_dir+'co_county.shp')\n", "\n", "#Generate queen weights\n", "w_queen=pysal.queen_from_shapefile(shp_dir+'co_county.shp')\n", "\n", "#Calculate minimum distance at which each county has at least one neighbor\n", "thresh=pysal.min_threshold_dist_from_shapefile(shp_dir+'co_county.shp')\n", "\n", "#Calculate binary distance band weights\n", "w_db_b=pysal.threshold_binaryW_from_shapefile(shp_dir+'co_county.shp',thresh)\n", "\n", "#Calculate continous distance band weights\n", "w_db_c=pysal.threshold_continuousW_from_shapefile(shp_dir+'co_county.shp',thresh)\n", "\n", "#Generate kernel based weights\n", "w_kern=pysal.kernelW_from_shapefile(shp_dir+'co_county.shp',function='gaussian')\n", "\n", "#Create list of weight matrices\n", "w_list=[w_rook,w_queen,w_db_b,w_db_c,w_kern]" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 36 }, { "cell_type": "markdown", "metadata": {}, "source": [ "Now we have the components required for an SLM, so we can define a function to facilitate the execution of multiple models. First, however, we must deal with Broomfield County. The county did not have assessment nor demographic profile information prior to 2001. This has led to the existence of missing data that are confounding the matrix shape requirements for this estimator. This is a less than ideal solution, but we are replacing missing data with within county means for each specific column." ] }, { "cell_type": "code", "collapsed": false, "input": [ "for col in co_TEL95.ix['Broomfield'].columns:\n", " co_TEL95.ix['Broomfield'][col].replace(NaN,co_TEL95.ix['Broomfield'][col].mean(),inplace=True)" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 37 }, { "cell_type": "markdown", "metadata": {}, "source": [ "Now we can develop the SLM function. Recall that its cross-sectional nature renders Gross State Product invariant. Thus, it must be removed." ] }, { "cell_type": "code", "collapsed": false, "input": [ "def slm_yr(df,df_name,yvar,x_list,yr,w,lag,null_check):\n", " \"\"\"This function is utility to facilitate looping through all weighting schemes.\n", " Note: data are subset to provide only non-null values in relative annual payroll per capita.\n", " The collection of annual payroll data in one county did not begin until 2002.\"\"\"\n", " #Set dependent variable\n", " #print '\\nSETTING DEPENDENT:',yvar\n", " y=df[df[null_check].notnull()].xs(yr,level='AUDIT_YEAR')[yvar].values\n", " \n", " #Adhere to the structure constraints imposed by PySAL\n", " y.shape=(len(y),1)\n", " \n", " #Create container for regressors\n", " X=[]\n", " \n", " #Add regressors to container in turn\n", " #print '\\nBUILDING REGRESSOR MATRIX'\n", " for reg in x_list:\n", "# print '\\t',reg\n", " X.append(df[df[null_check].notnull()].xs(yr,level='AUDIT_YEAR')[reg].values)\n", " \n", " #Check for null values\n", " #print DataFrame(X)\n", " #null_series=DataFrame(X).groupby(DataFrame(X).index).apply(count_missing)\n", " #print null_series\n", " #print null_series[null_series>0]\n", " \n", " #Convert from j x n to n x j (again to adhere to PySAL structure constraints\n", " #print '\\nTRANSPOSING REGRESSOR MATRIX'\n", " X=np.array(X).T\n", " \n", " #Row standardize weight matrix\n", " w.transform='r'\n", " \n", " #Run model\n", " slm=pysal.spreg.twosls_sp.GM_Lag(y,X,w=w,w_lags=lag,name_y=yvar,name_x=x_list,name_ds=df_name)\n", " return slm\n", "\n", "#Define subset of cross-sectionally varying columns\n", "cross_cty_econ=sorted(list(set(X_econ) - set(['gsp','st_unempr'])))\n", "cross_cty_rev=sorted(list(set(X_rev) - set(['gsp','st_unempr'])))\n", "\n", "slm_econ95=slm_yr(co_TEL95,'co_TEL',y_econ,cross_cty_econ[::-1],1995,w_list[0],2,null_var)\n", "\n", "print slm_econ95.summary" ], "language": "python", "metadata": {}, "outputs": [ { "output_type": "stream", "stream": "stdout", "text": [ "REGRESSION\n", "----------\n", "SUMMARY OF OUTPUT: SPATIAL TWO STAGE LEAST SQUARES\n", "--------------------------------------------------\n", "Data set : co_TEL\n", "Weights matrix : unknown\n", "Dependent Variable : pcap Number of Observations: 64\n", "Mean dependent var : 6576.1607 Number of Variables : 6\n", "S.D. dependent var : 5978.4921 Degrees of Freedom : 58\n", "Pseudo R-squared : 0.1700\n", "Spatial Pseudo R-squared: 0.1055\n", "\n", "------------------------------------------------------------------------------------\n", " Variable Coefficient Std.Error z-Statistic Probability\n", "------------------------------------------------------------------------------------\n", " CONSTANT 2003.6322092 2526.3963685 0.7930791 0.4277317\n", " W_pcap 0.3410880 0.3436879 0.9924353 0.3209852\n", " intensity_stock 900.9224267 12727.1264034 0.0707876 0.9435668\n", " permit_rate -11482.2800919 45917.1551509 -0.2500651 0.8025370\n", " prop_ratio 4673.8028903 2118.7610247 2.2059132 0.0273901\n", " vac_rate -2.0943183 34.8929105 -0.0600213 0.9521387\n", "------------------------------------------------------------------------------------\n", "Instrumented: W_pcap\n", "Instruments: W2_intensity_stock, W2_permit_rate, W2_prop_ratio, W2_vac_rate,\n", " W_intensity_stock, W_permit_rate, W_prop_ratio, W_vac_rate\n", "================================ END OF REPORT =====================================\n" ] } ], "prompt_number": 38 }, { "cell_type": "markdown", "metadata": {}, "source": [ "With the specification modified to accommodate cross-sectional invariance in statewide indicators, we can attempt to extract the summary information into a DataFrame (much like the OLS summary process above)." ] }, { "cell_type": "code", "collapsed": false, "input": [ "def slm_df(model,model_name,yr):\n", " print '***RUNNING',model_name,'***\\n'\n", " \n", " #Create containers for column names, estimates and p-values\n", " col_list=[]\n", " est_list=[]\n", " pval_list=[]\n", " \n", " #Populate estimate and p-value containers\n", " for col in model.name_x:\n", " col_list.append(col)\n", " for est in model.betas.tolist():\n", " est_list.append(est[0])\n", " for zstat_tuple in model.z_stat:\n", " pval_list.append(zstat_tuple[1])\n", " \n", " #Generate model DF (note that while W does not show up in the column names, it is in the estimates/pvals)\n", " slm_df=DataFrame({'var':col_list[1:],\n", " 'est':est_list[1:-1],\n", " 'pval':pval_list[1:-1]})\n", " \n", " #Include R^2\n", " slm_df['pr2']=model.pr2\n", " \n", " #Include model name\n", " slm_df['model']=model_name\n", " \n", " #Include model year\n", " slm_df['year']=yr\n", " \n", " return slm_df\n", "\n", "slm_df(slm_econ95,'slm_econ95',1995)" ], "language": "python", "metadata": {}, "outputs": [ { "output_type": "stream", "stream": "stdout", "text": [ "***RUNNING slm_econ95 ***\n", "\n" ] }, { "html": [ "

	est	pval	var	pr2	model	year
0	-2.094318	0.952139	vac_rate	0.169998	slm_econ95	1995
1	4673.802890	0.027390	prop_ratio	0.169998	slm_econ95	1995
2	-11482.280092	0.802537	permit_rate	0.169998	slm_econ95	1995
3	900.922427	0.943567	intensity_stock	0.169998	slm_econ95	1995

\n", "

" ], "metadata": {}, "output_type": "pyout", "prompt_number": 39, "text": [ " est pval var pr2 model year\n", "0 -2.094318 0.952139 vac_rate 0.169998 slm_econ95 1995\n", "1 4673.802890 0.027390 prop_ratio 0.169998 slm_econ95 1995\n", "2 -11482.280092 0.802537 permit_rate 0.169998 slm_econ95 1995\n", "3 900.922427 0.943567 intensity_stock 0.169998 slm_econ95 1995" ] } ], "prompt_number": 39 }, { "cell_type": "markdown", "metadata": {}, "source": [ "Now we can execute this model for each year, within each model specification." ] }, { "cell_type": "code", "collapsed": false, "input": [ "slm_list=[]\n", "for year in range(1995,2010):\n", " slm_econ=slm_yr(co_TEL95,'co_TEL',y_econ,cross_cty_econ[::-1],year,w_list[0],2,null_var)\n", " slm_rev=slm_yr(co_TEL95,'co_TEL',y_rev,cross_cty_rev[::-1],year,w_list[0],2,null_var)\n", " slm_list.append(slm_df(slm_econ,'slm_econ',year))\n", " slm_list.append(slm_df(slm_rev,'slm_rev',year))\n", " \n", "print slm_list[5]" ], "language": "python", "metadata": {}, "outputs": [ { "output_type": "stream", "stream": "stdout", "text": [ "***RUNNING slm_econ ***\n", "\n", "***RUNNING slm_rev ***\n", "\n", "***RUNNING slm_econ ***\n", "\n", "***RUNNING slm_rev ***\n", "\n", "***RUNNING slm_econ ***\n", "\n", "***RUNNING slm_rev ***\n", "\n", "***RUNNING slm_econ ***\n", "\n", "***RUNNING slm_rev ***\n", "\n", "***RUNNING" ] }, { "output_type": "stream", "stream": "stdout", "text": [ " slm_econ ***\n", "\n", "***RUNNING slm_rev ***\n", "\n", "***RUNNING slm_econ ***\n", "\n", "***RUNNING slm_rev ***\n", "\n", "***RUNNING slm_econ ***\n", "\n", "***RUNNING slm_rev ***\n", "\n", "***RUNNING slm_econ ***\n", "\n", "***RUNNING slm_rev ***\n", "\n", "***RUNNING" ] }, { "output_type": "stream", "stream": "stdout", "text": [ " slm_econ ***\n", "\n", "***RUNNING slm_rev ***\n", "\n", "***RUNNING slm_econ ***\n", "\n", "***RUNNING slm_rev ***\n", "\n", "***RUNNING slm_econ ***\n", "\n", "***RUNNING slm_rev ***\n", "\n", "***RUNNING slm_econ ***\n", "\n", "***RUNNING slm_rev ***\n", "\n", "***RUNNING" ] }, { "output_type": "stream", "stream": "stdout", "text": [ " slm_econ ***\n", "\n", "***RUNNING slm_rev ***\n", "\n", "***RUNNING slm_econ ***\n", "\n", "***RUNNING slm_rev ***\n", "\n", "***RUNNING slm_econ ***\n", "\n", "***RUNNING slm_rev ***\n", "\n", " est pval var pr2 model year\n", "0 26.739097 1.880800e-21 vac_rate 0.662894 slm_rev 1997\n", "1 -315.674329 7.564981e-02 prop_ratio 0.662894 slm_rev 1997\n", "2 -8506.077737 2.733771e-02 permit_rate 0.662894 slm_rev 1997\n", "3 0.025877 5.370525e-03 pcap 0.662894 slm_rev 1997\n", "4 1720.029785 5.561948e-01 lpop_growth 0.662894 slm_rev 1997\n", "5 -906.563954 3.380347e-02 intensity_stock 0.662894 slm_rev 1997\n" ] } ], "prompt_number": 40 }, { "cell_type": "markdown", "metadata": {}, "source": [ "The same must be done for each clustering variable... (or should it?)" ] }, { "cell_type": "code", "collapsed": false, "input": [ "# ##Capture all models for revenue clustering\n", "# #For each year...\n", "# for year in range(1995,2010):\n", "# print year\n", "# #... and each clustering dependent variable within year...\n", "# for i in range(len(y_clust_rev)):\n", "# print y_clust_rev[i]\n", "# #Run SLM and capture summary info\n", "# slm_clust_rev=slm_yr(co_TEL95,'co_TEL',y_clust_rev[i],cross_cty_rev,year,w_list[0],2,null_var)\n", "# slm_clust_econ=slm_yr(co_TEL95,'co_TEL',y_clust_econ[i],cross_cty_econ,year,w_list[0],2,null_var)\n", "# #Dump summary info in the DF list\n", "# slm_list.append(slm_df(slm_clust_rev,'slm_clust_rev'+str(year),year))\n", "# slm_list.append(slm_df(slm_clust_econ,'slm_clust_econ'+str(year),year))\n", " \n", "# #Concatenate all sets together\n", "# slm_results=pd.concat(slm_list).set_index(['var','model']).sortlevel(0)" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 41 }, { "cell_type": "code", "collapsed": false, "input": [ "#Concatenate all sets together\n", "slm_results=pd.concat(slm_list)\n", "\n", "#Integrate state unemployment rate\n", "slm_results['st_unempr']=slm_results['year'].map(state_unemp_dict)\n", "\n", "slm_results.head(20)" ], "language": "python", "metadata": {}, "outputs": [ { "html": [ "

	est	pval	var	pr2	model	year	st_unempr
0	-2.094318	9.521387e-01	vac_rate	0.169998	slm_econ	1995	4.0
1	4673.802890	2.739008e-02	prop_ratio	0.169998	slm_econ	1995	4.0
2	-11482.280092	8.025370e-01	permit_rate	0.169998	slm_econ	1995	4.0
3	900.922427	9.435668e-01	intensity_stock	0.169998	slm_econ	1995	4.0
0	25.799060	1.135082e-18	vac_rate	0.674969	slm_rev	1995	4.0
1	-470.055251	4.747703e-03	prop_ratio	0.674969	slm_rev	1995	4.0
2	-6364.202976	7.586142e-02	permit_rate	0.674969	slm_rev	1995	4.0
3	0.027099	2.715553e-03	pcap	0.674969	slm_rev	1995	4.0
4	1312.640704	3.896742e-01	lpop_growth	0.674969	slm_rev	1995	4.0
5	-325.479324	7.241664e-01	intensity_stock	0.674969	slm_rev	1995	4.0
0	-16.348986	6.502216e-01	vac_rate	0.232434	slm_econ	1996	4.2
1	4398.160680	2.369744e-02	prop_ratio	0.232434	slm_econ	1996	4.2
2	47999.830607	2.037459e-01	permit_rate	0.232434	slm_econ	1996	4.2
3	-16310.527203	1.245665e-01	intensity_stock	0.232434	slm_econ	1996	4.2
0	27.687441	6.588370e-20	vac_rate	0.644088	slm_rev	1996	4.2
1	-471.685092	4.933895e-03	prop_ratio	0.644088	slm_rev	1996	4.2
2	2412.836247	5.163525e-01	permit_rate	0.644088	slm_rev	1996	4.2
3	0.022416	3.125443e-02	pcap	0.644088	slm_rev	1996	4.2
4	-3539.952387	1.516840e-01	lpop_growth	0.644088	slm_rev	1996	4.2
5	-856.872113	3.323542e-01	intensity_stock	0.644088	slm_rev	1996	4.2

\n", "

" ], "metadata": {}, "output_type": "pyout", "prompt_number": 42, "text": [ " est pval var pr2 model year st_unempr\n", "0 -2.094318 9.521387e-01 vac_rate 0.169998 slm_econ 1995 4.0\n", "1 4673.802890 2.739008e-02 prop_ratio 0.169998 slm_econ 1995 4.0\n", "2 -11482.280092 8.025370e-01 permit_rate 0.169998 slm_econ 1995 4.0\n", "3 900.922427 9.435668e-01 intensity_stock 0.169998 slm_econ 1995 4.0\n", "0 25.799060 1.135082e-18 vac_rate 0.674969 slm_rev 1995 4.0\n", "1 -470.055251 4.747703e-03 prop_ratio 0.674969 slm_rev 1995 4.0\n", "2 -6364.202976 7.586142e-02 permit_rate 0.674969 slm_rev 1995 4.0\n", "3 0.027099 2.715553e-03 pcap 0.674969 slm_rev 1995 4.0\n", "4 1312.640704 3.896742e-01 lpop_growth 0.674969 slm_rev 1995 4.0\n", "5 -325.479324 7.241664e-01 intensity_stock 0.674969 slm_rev 1995 4.0\n", "0 -16.348986 6.502216e-01 vac_rate 0.232434 slm_econ 1996 4.2\n", "1 4398.160680 2.369744e-02 prop_ratio 0.232434 slm_econ 1996 4.2\n", "2 47999.830607 2.037459e-01 permit_rate 0.232434 slm_econ 1996 4.2\n", "3 -16310.527203 1.245665e-01 intensity_stock 0.232434 slm_econ 1996 4.2\n", "0 27.687441 6.588370e-20 vac_rate 0.644088 slm_rev 1996 4.2\n", "1 -471.685092 4.933895e-03 prop_ratio 0.644088 slm_rev 1996 4.2\n", "2 2412.836247 5.163525e-01 permit_rate 0.644088 slm_rev 1996 4.2\n", "3 0.022416 3.125443e-02 pcap 0.644088 slm_rev 1996 4.2\n", "4 -3539.952387 1.516840e-01 lpop_growth 0.644088 slm_rev 1996 4.2\n", "5 -856.872113 3.323542e-01 intensity_stock 0.644088 slm_rev 1996 4.2" ] } ], "prompt_number": 42 }, { "cell_type": "markdown", "metadata": {}, "source": [ "At this point, we should really be proceeding with the Spatial Panel Model. This, however, requires some manipulation in R because such an estimator is not available in Python (at least not in the PySAL package, nor any other package that I have seen to date). In the interests of time, at present we proceed to exploring the results of the OLS and SLM models." ] }, { "cell_type": "heading", "level": 3, "metadata": {}, "source": [ "Visualizing Estimation Results" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "As we did in *CO_TEL_convergence*, we will visualize the data in R because ggplot2 provides an easier method of value-based coloring." ] }, { "cell_type": "code", "collapsed": false, "input": [ "#%load_ext rmagic" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 43 }, { "cell_type": "code", "collapsed": false, "input": [ "ols_results.to_csv(workdir+'tcr_ols_results.csv')\n", "slm_results.to_csv(workdir+'tcr_slm_results.csv')" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 44 }, { "cell_type": "code", "collapsed": false, "input": [ "HTML(slm_results.to_html())" ], "language": "python", "metadata": {}, "outputs": [ { "html": [ "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "

	est	pval	var	pr2	model	year	st_unempr
0	-2.094318	9.521387e-01	vac_rate	0.169998	slm_econ	1995	4.0
1	4673.802890	2.739008e-02	prop_ratio	0.169998	slm_econ	1995	4.0
2	-11482.280092	8.025370e-01	permit_rate	0.169998	slm_econ	1995	4.0
3	900.922427	9.435668e-01	intensity_stock	0.169998	slm_econ	1995	4.0
0	25.799060	1.135082e-18	vac_rate	0.674969	slm_rev	1995	4.0
1	-470.055251	4.747703e-03	prop_ratio	0.674969	slm_rev	1995	4.0
2	-6364.202976	7.586142e-02	permit_rate	0.674969	slm_rev	1995	4.0
3	0.027099	2.715553e-03	pcap	0.674969	slm_rev	1995	4.0
4	1312.640704	3.896742e-01	lpop_growth	0.674969	slm_rev	1995	4.0
5	-325.479324	7.241664e-01	intensity_stock	0.674969	slm_rev	1995	4.0
0	-16.348986	6.502216e-01	vac_rate	0.232434	slm_econ	1996	4.2
1	4398.160680	2.369744e-02	prop_ratio	0.232434	slm_econ	1996	4.2
2	47999.830607	2.037459e-01	permit_rate	0.232434	slm_econ	1996	4.2
3	-16310.527203	1.245665e-01	intensity_stock	0.232434	slm_econ	1996	4.2
0	27.687441	6.588370e-20	vac_rate	0.644088	slm_rev	1996	4.2
1	-471.685092	4.933895e-03	prop_ratio	0.644088	slm_rev	1996	4.2
2	2412.836247	5.163525e-01	permit_rate	0.644088	slm_rev	1996	4.2
3	0.022416	3.125443e-02	pcap	0.644088	slm_rev	1996	4.2
4	-3539.952387	1.516840e-01	lpop_growth	0.644088	slm_rev	1996	4.2
5	-856.872113	3.323542e-01	intensity_stock	0.644088	slm_rev	1996	4.2
0	-4.847488	8.976131e-01	vac_rate	0.195421	slm_econ	1997	3.4
1	5945.379340	1.177646e-02	prop_ratio	0.195421	slm_econ	1997	3.4
2	-2201.309246	9.565427e-01	permit_rate	0.195421	slm_econ	1997	3.4
3	-3375.376504	5.706255e-01	intensity_stock	0.195421	slm_econ	1997	3.4
0	26.739097	1.880800e-21	vac_rate	0.662894	slm_rev	1997	3.4
1	-315.674329	7.564981e-02	prop_ratio	0.662894	slm_rev	1997	3.4
2	-8506.077737	2.733771e-02	permit_rate	0.662894	slm_rev	1997	3.4
3	0.025877	5.370525e-03	pcap	0.662894	slm_rev	1997	3.4
4	1720.029785	5.561948e-01	lpop_growth	0.662894	slm_rev	1997	3.4
5	-906.563954	3.380347e-02	intensity_stock	0.662894	slm_rev	1997	3.4
0	-13.564872	7.282480e-01	vac_rate	0.194247	slm_econ	1998	3.5
1	5129.452505	3.005339e-02	prop_ratio	0.194247	slm_econ	1998	3.5
2	26449.073535	5.632933e-01	permit_rate	0.194247	slm_econ	1998	3.5
3	-2141.557551	6.966523e-01	intensity_stock	0.194247	slm_econ	1998	3.5
0	28.229999	1.460064e-24	vac_rate	0.690899	slm_rev	1998	3.5
1	-294.848446	8.441574e-02	prop_ratio	0.690899	slm_rev	1998	3.5
2	-16380.003976	7.190607e-05	permit_rate	0.690899	slm_rev	1998	3.5
3	0.034774	6.842840e-05	pcap	0.690899	slm_rev	1998	3.5
4	7726.201094	1.637476e-02	lpop_growth	0.690899	slm_rev	1998	3.5
5	-924.722973	1.504198e-02	intensity_stock	0.690899	slm_rev	1998	3.5
0	-20.774152	6.149559e-01	vac_rate	0.187823	slm_econ	1999	3.0
1	5228.185252	4.599714e-02	prop_ratio	0.187823	slm_econ	1999	3.0
2	23931.142021	6.400282e-01	permit_rate	0.187823	slm_econ	1999	3.0
3	1918.626166	6.889240e-01	intensity_stock	0.187823	slm_econ	1999	3.0
0	26.570657	1.994488e-19	vac_rate	0.654169	slm_rev	1999	3.0
1	-360.689060	6.275398e-02	prop_ratio	0.654169	slm_rev	1999	3.0
2	-10146.453798	4.189333e-02	permit_rate	0.654169	slm_rev	1999	3.0
3	0.033114	9.021667e-05	pcap	0.654169	slm_rev	1999	3.0
4	2412.760156	4.553753e-01	lpop_growth	0.654169	slm_rev	1999	3.0
5	-839.428685	2.185437e-02	intensity_stock	0.654169	slm_rev	1999	3.0
0	-61.727656	2.220864e-01	vac_rate	0.192146	slm_econ	2000	2.7
1	4686.192411	7.233134e-02	prop_ratio	0.192146	slm_econ	2000	2.7
2	190919.536098	4.156700e-01	permit_rate	0.192146	slm_econ	2000	2.7
3	3527.993508	4.836277e-01	intensity_stock	0.192146	slm_econ	2000	2.7
0	32.870589	5.989494e-20	vac_rate	0.636441	slm_rev	2000	2.7
1	-471.126354	1.222936e-02	prop_ratio	0.636441	slm_rev	2000	2.7
2	-43917.706013	2.140134e-02	permit_rate	0.636441	slm_rev	2000	2.7
3	0.041066	1.454352e-06	pcap	0.636441	slm_rev	2000	2.7
4	2259.479169	3.246444e-01	lpop_growth	0.636441	slm_rev	2000	2.7
5	-954.814102	8.325711e-03	intensity_stock	0.636441	slm_rev	2000	2.7
0	-77.554834	1.175968e-01	vac_rate	0.233445	slm_econ	2001	3.8
1	4480.404240	6.020616e-02	prop_ratio	0.233445	slm_econ	2001	3.8
2	48342.027878	4.280830e-01	permit_rate	0.233445	slm_econ	2001	3.8
3	5246.644817	1.522870e-01	intensity_stock	0.233445	slm_econ	2001	3.8
0	34.449089	3.093784e-19	vac_rate	0.632418	slm_rev	2001	3.8
1	-332.256089	6.950290e-02	prop_ratio	0.632418	slm_rev	2001	3.8
2	-9398.972323	3.305903e-02	permit_rate	0.632418	slm_rev	2001	3.8
3	0.048124	4.462320e-08	pcap	0.632418	slm_rev	2001	3.8
4	-1949.801790	8.341674e-02	lpop_growth	0.632418	slm_rev	2001	3.8
5	-620.937210	1.903018e-02	intensity_stock	0.632418	slm_rev	2001	3.8
0	-54.580487	2.367983e-01	vac_rate	0.273145	slm_econ	2002	5.7
1	2254.118066	2.633969e-01	prop_ratio	0.273145	slm_econ	2002	5.7
2	122181.401953	2.575673e-02	permit_rate	0.273145	slm_econ	2002	5.7
3	5049.600815	1.415358e-01	intensity_stock	0.273145	slm_econ	2002	5.7
0	36.037854	1.581346e-15	vac_rate	0.576833	slm_rev	2002	5.7
1	-648.293666	2.695821e-03	prop_ratio	0.576833	slm_rev	2002	5.7
2	-5862.859870	3.325828e-01	permit_rate	0.576833	slm_rev	2002	5.7
3	0.051539	4.544804e-06	pcap	0.576833	slm_rev	2002	5.7
4	366.571141	9.057923e-01	lpop_growth	0.576833	slm_rev	2002	5.7
5	-511.864063	1.336298e-01	intensity_stock	0.576833	slm_rev	2002	5.7
0	-38.480254	4.244922e-01	vac_rate	0.211340	slm_econ	2003	6.1
1	3482.690915	1.077937e-01	prop_ratio	0.211340	slm_econ	2003	6.1
2	67379.794550	2.821857e-01	permit_rate	0.211340	slm_econ	2003	6.1
3	4017.259505	2.456099e-01	intensity_stock	0.211340	slm_econ	2003	6.1
0	33.293603	2.076206e-19	vac_rate	0.631206	slm_rev	2003	6.1
1	-688.676569	9.023839e-05	prop_ratio	0.631206	slm_rev	2003	6.1
2	-1004.364722	8.689395e-01	permit_rate	0.631206	slm_rev	2003	6.1
3	0.046707	4.419849e-07	pcap	0.631206	slm_rev	2003	6.1
4	-3912.358647	4.022211e-01	lpop_growth	0.631206	slm_rev	2003	6.1
5	-264.332799	3.622043e-01	intensity_stock	0.631206	slm_rev	2003	6.1
0	-44.314071	3.592491e-01	vac_rate	0.223234	slm_econ	2004	5.6
1	3797.887900	4.736302e-02	prop_ratio	0.223234	slm_econ	2004	5.6
2	104579.654708	1.665196e-01	permit_rate	0.223234	slm_econ	2004	5.6
3	2888.504167	3.179798e-01	intensity_stock	0.223234	slm_econ	2004	5.6
0	34.564514	1.951334e-21	vac_rate	0.659799	slm_rev	2004	5.6
1	-582.846221	7.977103e-05	prop_ratio	0.659799	slm_rev	2004	5.6
2	-11400.063888	8.472239e-02	permit_rate	0.659799	slm_rev	2004	5.6
3	0.051253	6.602262e-09	pcap	0.659799	slm_rev	2004	5.6
4	72.490876	9.878567e-01	lpop_growth	0.659799	slm_rev	2004	5.6
5	-354.921267	1.053262e-01	intensity_stock	0.659799	slm_rev	2004	5.6
0	-40.026955	4.483235e-01	vac_rate	0.229168	slm_econ	2005	5.1
1	4802.586194	2.190209e-02	prop_ratio	0.229168	slm_econ	2005	5.1
2	82966.813330	3.040512e-01	permit_rate	0.229168	slm_econ	2005	5.1
3	2798.267141	2.932797e-01	intensity_stock	0.229168	slm_econ	2005	5.1
0	35.934092	2.623176e-19	vac_rate	0.645814	slm_rev	2005	5.1
1	-602.845815	2.278566e-04	prop_ratio	0.645814	slm_rev	2005	5.1
2	-15196.697652	6.437954e-02	permit_rate	0.645814	slm_rev	2005	5.1
3	0.048505	6.990021e-08	pcap	0.645814	slm_rev	2005	5.1
4	5546.380429	3.447653e-01	lpop_growth	0.645814	slm_rev	2005	5.1
5	-491.559409	1.807049e-02	intensity_stock	0.645814	slm_rev	2005	5.1
0	-52.893420	3.341184e-01	vac_rate	0.246212	slm_econ	2006	4.3
1	5429.234172	9.287807e-03	prop_ratio	0.246212	slm_econ	2006	4.3
2	111460.376741	1.900461e-01	permit_rate	0.246212	slm_econ	2006	4.3
3	2238.947189	3.663905e-01	intensity_stock	0.246212	slm_econ	2006	4.3
0	38.342519	5.813789e-14	vac_rate	0.636088	slm_rev	2006	4.3
1	-822.749991	1.999012e-06	prop_ratio	0.636088	slm_rev	2006	4.3
2	-6592.620241	4.655747e-01	permit_rate	0.636088	slm_rev	2006	4.3
3	0.050709	7.105195e-08	pcap	0.636088	slm_rev	2006	4.3
4	-89.176355	9.907440e-01	lpop_growth	0.636088	slm_rev	2006	4.3
5	-294.694629	1.456341e-01	intensity_stock	0.636088	slm_rev	2006	4.3
0	-64.867077	2.282071e-01	vac_rate	0.290308	slm_econ	2007	3.8
1	4536.095479	2.186696e-02	prop_ratio	0.290308	slm_econ	2007	3.8
2	190861.023223	3.310253e-02	permit_rate	0.290308	slm_econ	2007	3.8
3	2331.862427	2.920015e-01	intensity_stock	0.290308	slm_econ	2007	3.8
0	41.230236	1.286264e-16	vac_rate	0.609323	slm_rev	2007	3.8
1	-952.886172	5.120721e-07	prop_ratio	0.609323	slm_rev	2007	3.8
2	-9447.408867	3.070681e-01	permit_rate	0.609323	slm_rev	2007	3.8
3	0.058855	1.470501e-07	pcap	0.609323	slm_rev	2007	3.8
4	2040.281395	6.806850e-01	lpop_growth	0.609323	slm_rev	2007	3.8
5	-320.788378	1.360186e-01	intensity_stock	0.609323	slm_rev	2007	3.8
0	-103.704327	5.082072e-02	vac_rate	0.333334	slm_econ	2008	4.8
1	4023.097170	2.716576e-02	prop_ratio	0.333334	slm_econ	2008	4.8
2	349425.514454	8.410474e-04	permit_rate	0.333334	slm_econ	2008	4.8
3	1547.112856	4.796538e-01	intensity_stock	0.333334	slm_econ	2008	4.8
0	41.259332	6.454407e-14	vac_rate	0.576289	slm_rev	2008	4.8
1	-771.546230	3.634214e-05	prop_ratio	0.576289	slm_rev	2008	4.8
2	-3635.896454	8.024777e-01	permit_rate	0.576289	slm_rev	2008	4.8
3	0.055688	4.788989e-06	pcap	0.576289	slm_rev	2008	4.8
4	-2172.829689	6.790247e-01	lpop_growth	0.576289	slm_rev	2008	4.8
5	-478.562764	3.457945e-02	intensity_stock	0.576289	slm_rev	2008	4.8
0	-118.745684	7.184241e-03	vac_rate	0.391351	slm_econ	2009	8.1
1	3296.047495	2.519380e-02	prop_ratio	0.391351	slm_econ	2009	8.1
2	599436.856220	1.296274e-06	permit_rate	0.391351	slm_econ	2009	8.1
3	1085.613147	4.909187e-01	intensity_stock	0.391351	slm_econ	2009	8.1
0	44.214173	1.815731e-13	vac_rate	0.562230	slm_rev	2009	8.1
1	-796.193051	3.907967e-05	prop_ratio	0.562230	slm_rev	2009	8.1
2	-4981.305649	7.838575e-01	permit_rate	0.562230	slm_rev	2009	8.1
3	0.053500	5.659704e-04	pcap	0.562230	slm_rev	2009	8.1
4	-4389.554704	4.397923e-01	lpop_growth	0.562230	slm_rev	2009	8.1
5	-260.926417	2.027842e-01	intensity_stock	0.562230	slm_rev	2009	8.1

" ], "metadata": {}, "output_type": "pyout", "prompt_number": 45, "text": [ "" ] } ], "prompt_number": 45 }, { "cell_type": "code", "collapsed": false, "input": [], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 45 } ], "metadata": {} } ] }