-
What is the serving rate for each scenario throughout the day, comparing to the overall serving rate? Serving rate is the number of successfully matched trip in a period of time. For example, is the serving rate higher in the rush hour or at night?
-
For those trips that could not be served, do they follow a spatial or temporal pattern? For example, are most of those trips originated in particular regions, and of certain times?
-
We also limit vehicle capacity to at most 4 passengers. Were there vehicles violating this condition? If so, can you show any pattern about these vehicles? For example, how many of them were violating, and where were they distributed in both time and space?
Note : Steps for the creation of the dataset for visualization purposes are available in the supporting .ipynb notebooks which can be found in the repo.
Task 1: What is the serving rate for each scenario throughout the day, comparing to the overall serving rate? Serving rate is the number of successfully matched trip in a period of time. For example, is the serving rate higher in the rush hour or at night?
https://bl.ocks.org/vvt221/0820e350bd4019b50fc4f7b903c1435e
The above plot shows the hourly serving rate against the mean serving rate based on the number of requests made. Serving rate is very high in the morning hours accounting for everyone probably commuting to the workplaces. There is a dip around 11 am and stays below the mean till 8.00 pm from where it rises above the average serving rate till 10.00pm after which it again falls below the averge serving rate. Clearly, the serving rate is very high during the morning hours.
Task 2: For those trips that could not be served, do they follow a spatial or temporal pattern? For example, are most of those trips originated in particular regions, and of certain times?
https://bl.ocks.org/vvt221/7308a3becd0c42c40a22ff735f3fb555
From the plot it can be seen that a majority of the trips are not served in the early hours of the morning from 12.00 AM to 4.00 AM. However, this trend suddenly drops between 5.00 AM and 11.00 AM . This could be because this is the time duration around which people commute to their workplaces and they cannot afford to be late which is why the number of trips where the rides are dropped is low. There is once again an increase in the number of trips dropped. However there is a sharp decline in the number of trips dropped at 9.00 pm ,which could be due to people wanting to reach home as early as possible. The high percentage of ride dropping during the hours discussed above could be due to the high price being charged during these hours but data is not available to verify that.
https://bl.ocks.org/vvt221/361d67866f021d266f52fd930f7792bb
This plot shows the bar charts based on the counts of the Actual_Pickup station where most rides could not be served. The counts are then sorted in decreasing order of their frequency. Station ID 229 and Station ID 1196 hold the record for having the highest number of dropped records. Further exploration is done in the spatial plot below:
https://bl.ocks.org/vvt221/89551b051b549965b8ac67b375280bbd
This plot shows the number of dropped rides at census tract level in Manhattan. From the plot we can see that a large number of rides are dropped in the Central Park region. There are few census tracts near the financial district which hace a high drop rate. Another region with a high ride drop rate is the area in the upper regions of Manhattan.
Task 4: We also limit vehicle capacity to at most 4 passengers. Were there vehicles violating this condition? Ifso, can you show any pattern about these vehicles? For example, how many of them were violating,and where were they distributed in both time and space?
https://bl.ocks.org/vvt221/8bf02bece90b84e4d349cc51645e23d0
This plot shows the hourly distribution of counts of rides where the number of passengers exceeded 4. There are a lot of such rides during the late night hours and early morning hours. But the number is less during the morning and afternoon hours. It could be possible due to garbage values because the passenger count exceeds 10 and goes on till 68 which are probably garbage values.
https://bl.ocks.org/vvt221/dc8610c07a8ba2818e291a37eacc89f6
This plot shows the number of passengers exceeding 4 varies throughout the hour of the day and pretty much is consistent with the above plot but can be used to see how the distribution varies hourly at a granular level.
https://bl.ocks.org/vvt221/ca1a385328565e7df266ea28e258db10
This plot over here shows the spatial distribution of rides where the passenger count exceeds 4 at the census tract level. The preponderance of such rides all over Manhattan leads us to believe that it could possible be due to garbage values being entered in the passenger info field.