Analyze and visualize ride share data with Python and Matplotlib.
Use the Pandas Library to analyze rideshare data and visualize findings using the Matplotlib Library.
Data was stored in two separate csv files, so the files were read into a notebook and merged on city using an inner join with pd.merge() and stored in a data frame, combined_data.
I wanted to showcase the relationship between four variables:
and believed a bubble plot would showcase this best, with the total number of rides as the x-axis, average fare cost as the y-axis, city types noted by marker color, and the total number of drivers being represented by the marker size.
I found three different city types using .unique() on the type column of the combined data. I then grouped the combined data on both type and city, storing that object in a variable named grouped_data.
For the total number of rides per city, I used .count() on the ride_id of grouped_data and further filtered those values by city type. I repeated the same process for both average fare and driver count, but used .mean() instead on the fare and driver_count columns. Because the driver count for each unique city was the same for every entry, .mean() was used so that the values could be accessed without be changed.
All these values were then plotted in a scatter plot and formatted for color and labeling.
Next, I wanted to find the percent of total fares, rides, and drivers by city type, which would be best depicted in pie charts.
Total fares were calculated using .sum() on the fare column of grouped_data, first created for the bubble plot, and then separated by the three city types. The sum for each city type was stored in a variable and those variables were then plotted in pie chart.
Total rides were calculated using .value_counts() on the type column of the original combined_data data frame. No further manipulation was required, so the values were stored in an array to be plotted, with each city type being accessed through indexing.
~
The count of total drivers by city type has already been accessed in the bubble plot section and stored in a variable, so I used that same variable and took the sum of each city type to plot.