In this project we set off to study the tendencies on fares varying from different types of city such as Urban, Rural, and Suburban for ride shares, and the amount of drivers that each type of city uses.
We continue to practice our Pandas skills on this project, now creating graphs that can help us visualize the outcomes that we obtained from studying the data, with Matplotlib. This is an extensive library that works within Python's ecosystem.
As we navigate and make calculations throughout the data, we can recognize that there is key information hidden within all those seemingly random numbers. By obtaining the measures of central tendency, for example, we can compare the effectiveness of PyBer in the different types of cities. Consider the following findings: "There is one outlier in the urban ride count data. Also, the average number of rides in the rural cities is about 4- and 3.5 times lower per city than the urban and suburban cities, respectively" (Module 5.4.4). What does this mean? Even though there is one unusual data point in the urban cities, this does not affect the number of rides taken in the other two cities. As one might expect, the difference between the rural cities, urban and suburban cities when it comes down to ordering a ride-share service is significant; however, we now have the statistics to prove this notion true. (See boxplot below).
Furthermore, it is important to keep in mind that having less people order a ride through PyBer in a rural city does not mean that the fares would be cheaper. On the contrary, the fare prices go up significantly for those who decide to take a ride in a rural city in comparison to their fellow riders in suburban and urban cities. And this makes sense, drivers in rural cities get less rides than a driver in an urban city who may be requested one ride after the other. But PyBer makes up for this difference by having higher average fare prices in the rural cities. (See boxplot below).
We also calculated the average number of drivers per city type and the outcome is as expected. The higher demand in the urban cities attracts a lot more drivers to work there. The difference in drivers by city type is notorious and we can reiterate our previous analysis. (See boxplot below).
Finally, diving deeper into the analysis of the performance of PyBer in ride-sharing across the different types of cities, we can consider the following graph:
With the information we have now about the tendencies of ride-share in these three types of cities, what can PyBer do to better its performance and create more revenue in all three of them? When looking at the peak points in this line graph, we see that all three of the types of city increase the total fares during the springtime; in between February and March, with the urban cities constantly rising throughout the month when comparing them with suburban and rural cities. Rural cities peak on demand during the month of April but continue in a consistent line through May; while the urban cities see a notorious decline during this time. There could be a number of factors that affect this decline that is also observed in suburban cities: the end of a school semester, people leaving for a summer vacation, etc.
We can say that PyBer does not need to gain a higher number of drivers in rural cities, as they have a steady total of fares that carry higher prices. And though the prices are lower in urban cities, this does not prevent the total fares from declining when preparing for the second quarter of the year. When these statistics show the relationship that urban, suburban, and rural cities have, we must put first the Law of Supply. An economic principle that has been a guide to free markets, we can rely that it will be a successful strategy as there are not many outliers that can interfere with the conditions for pricing and availability.