deletions | additions       diff --git a/Extracting Data From Figures.md b/Extracting Data From Figures.md  index 09f72a9..29c92e3 100644  --- a/Extracting Data From Figures.md  +++ b/Extracting Data From Figures.md 
...
## Extracting information from tables and graphs   See [related question on Stats.stackexchange](http://stats.stackexchange.com/a/14440/1381) To extract information from a figure, the general method is:  1. Identify upload an image   2. set  the data that is associated with x and y scales by indicating the values at two points on  each treatment axis   3. indicate if the scale is linear, log, etc,   4. click on the points.  *note:* If the experiment has many factors, the paper may not report the mean Some software programs automatically recognize lines or points. I am usually after points,  and statistics for each treatment. Often, the reported data will reflect in practice  the results are often too inconsistent to be helpful even with 100s  of more than points. I have not found  one treatment (for example, if there was no effect of the treatment on the quantity of interest). In some cases it will that distinguishes different symbols. This feature could  be possible to obtain worth  the values trouble  for each treatment, e.g. if there are _n-1_ values and _n_ treatments. If digitizing lines, but  this is not the case, the treatment names and definitions should be changed to indicate the data reflect the results of more than one experimental treatment. commonly required.  2. Enter The program returns each point as an x-y matrix.  Often it helps selecting points if  the mean value image is zoomed, either by uploading a zoomed version  of the trait image or using the zooming feature available in some of the programs.  ### List of available Programs  I have experience with the following programs. All of these work well fine. Except in contexts where measurement error is very small, error from graph scraping is insignificant (e.g. error from digitization << size of error bars or uncertainty in the estimate). I have not tested the accuracy of any of these programs, but it would be interesting to compare among users, among programs, and against the results of reproduced statistical analyses.  3. Enter the `statname`, `stat`, and number of replicates, `n`  associated with the mean  * `stat` is the value of the `statname` (i.e. `statname` might be  ’standard deviation’ (SD) and the `stat` is the numerical value  of the statistic)  * Always measure size of error bar from the mean to the end of an  error bar. This is the value when presented as ( _X_ ± _SE_) or  _X(SE)_ and may be found in a table or on a graph.  * Sometimes CI and LSD are presented as the entire range from the  lower to the upper end of the confidence interval. In this case,  take 1/2 of the interval representing the distance from the mean  to the upper or lower bound  * [Digitizer](http://digitizer.sourceforge.net/) (shareware) auto point / line recognition. Available in Ubuntu repository (engauge-digitizer)  * [Get Data](http://www.getdata-graph-digitizer.com/) (shareware) has zoom window, auto point / line recognition  * [DigitizeIt](http://www.digitizeit.de/) (shareware) auto point / line recognition  * [ImageJ](http://rsbweb.nih.gov/ij/) (open source, most extensible after R digitize)  * [R digitize](http://cran.r-project.org/web/packages/digitize/index.html) (free, open source), because it simplifies the processs of getting data from the graph into an analysis by keeping all of the steps in R. See the [tutorial in R-Journal](http://journal.r-project.org/archive/2011-1/RJournal_2011-1_Poisot.pdf)  * [GrabIt!](http://www.datatrendsoftware.com/home.html) (free demo, $69) Excel   plug-in   I have not used these:  * [WebPlotDigitzer](http://arohatgi.info/WebPlotDigitizer/app/) (free, online). Extracts data from images. Demo [here](http://blog.plot.ly/post/70293893434/automatically-grab-data-from-an-image-with).   * [GraphClick](http://www.arizona-software.ch/graphclick/) (Mac, $8)  * [g3data](http://www.frantz.fi/software/g3data.php) (open source - GNU GPL) Has zoom window, no auto-recognition. Available in Ubuntu repository.  See [related question on Stats.stackexchange](http://stats.stackexchange.com/a/14440/1381)  1. Identify the data that is associated with each treatment  *note:* If the experiment has many factors, the paper may not report the mean and statistics for each treatment. Often, the reported data will reflect the results of more than one treatment (for example, if there was no effect of the treatment on the quantity of interest). In some cases it will be possible to obtain the values for each treatment, e.g. if there are _n-1_ values and _n_ treatments. If this is not the case, the treatment names and definitions should be changed to indicate the data reflect the results of more than one experimental treatment.   2. Enter the mean value of the trait  3. Enter the `statname`, `stat`, and number of replicates, `n` associated with the mean  * `stat` is the value of the `statname` (i.e. `statname` might be ’standard deviation’ (SD) and the `stat` is the numerical value of the statistic)  * Always measure size of error bar from the mean to the end of an error bar. This is the value when presented as ( _X_ ± _SE_) or _X(SE)_ and may be found in a table or on a graph.  * Sometimes CI and LSD are presented as the entire range from the lower to the upper end of the confidence interval. In this case, take 1/2 of the interval representing the distance from the mean to the upper or lower bound.  ### Extracting Data using R 
...
5. Upload csv to the project file in google spreadsheet, or open as  excel/openoffice and copy/paste to google spreadsheet    ### Extracting Data From a Figure using GetData  1. Open PDF in Adobe Reader.