This is a article about CitiBike Sharing Project in PUI2016 class. The Report includes four parts Abstract, Data, Analysis and Result, which will be described in the following parts.

The idea of citibike sharing analysis is inspired by Professor Federica's lecture analysis. I want to discover the relationship between subscriber and commuter. The question is: Does people without membership really less likely than Subscripter to commute by bike? As a result, the null hypothesis is proposed as - The ratio of Subscripter biking on week days to Subscripter biking on weekends is the same or lower than the ratio of Customer biking on week days to Customer biking on weekends, which can be formulated as follow:

\begin{equation}
H_{0}:\frac{S_{weekend}}{S_{week}}\leq\frac{C_{weekend}}{C_{week}}\nonumber \\
\end{equation}

\begin{equation}
H_{1}:\frac{S_{weekend}}{S_{week}}>\frac{C_{weekend}}{C_{week}}\nonumber \\
\end{equation}

significance level:

\begin{equation} \alpha=0.05\nonumber \\ \end{equation}In this analysis, the data is based on the citibike usage in Jan. 2015. There are several attributes in dataset. For example , tripduration, start time, stop time, start station id, date ... etc. However, "usertype" and "date" are the only two attributes required in the analysis. Therefore, the selected dataframe is shown in Fig.1

Henry Linover 1 year ago · PublicI cant edit the text caption, it should be Fig 1: Dataframe for the analysis