Table 1.
Data Fields of Taxi Traces.
Fig 1.
A route plotted by a taxi’s trace data.
Each point denotes one reported GPS data with longitude and latitude information. The line segment between two consecutive points is approximately considered to be the actual driving route of the taxi. Note that di is the distance between two consecutive points calculated by above Formula (1) in this subsection.
Fig 2.
Daily Net Income comparison for different gas consumption = 6,7,8 liter/100km.
Fig 3.
The yellow periods of time denote the taxi’s working time, and we only take the time period in one day into account, so the part after t3 that passes 24 o’clock will not be cumulated into that day’s working time.
Fig 4.
The probability density distributions of taxis income in Beijing during Nov.- Dec. 2012.
(a) The magenta curve illustrates the net income of all 8846 taxis, and the blue square curve illustrates that of the selected 3590 taxis which operate every day during two months. (b) The number of selected taxis in Low Income, Medium Low Income, Medium High Income and High Income are 525, 542, 510, and 516, respectively. (c) The average daily income of taxis in each group are, 532.387, 360.985, 139.124, and 51.7747 CNY, respectively. (d) The shift from Net Income to Gross Income (The red diamond curve) denotes the taxi daily cost, which is inferred to be 337.5 CNY for each Taxi.
Fig 5.
(a) This proportion statistics are based on the selected 3,590 taxis, among which, four income level taxi drivers are selected. (b) The proportions are taxi number, passenger number, of four income levels of taxi total number of taxis, total number of passengers served, total number of grids where drivers pick up and drop off the passengers. (c) The pick-up and drop-off grids covered by all these four income level taxis occupy about 41 and 54 percentage of all city grids, respectively.
Fig 6.
The average daily working time and served passengers of four groups of selected 3590 taxis.
Fig 7.
The probability distributions of the distance per trip.
Fig 8.
The probability distributions of the distance per trip shorter than 100km.
Fig 9.
The pick-up and drop-off diversity of different income groups of taxis.
Fig 10.
The number of picking up grids with respect to the pick-up diversity
of each taxi. Four clusters correspond to four income level taxis.
Fig 11.
The number of dropping off grids with respect to the drop-off diversity
of each taxi.
Fig 12.
The number of dropping off grids with respect to the diversity of each grid
, the black crosses represent each grid where pick-ups happen,totaling 12176, while the red asterisks represent the grids where high income taxis pick up more than 500 times, approximately 1 time per taxi, totaling 508.
Fig 13.
The drop-off distributions of high and non-high income taxis, the forepart is zoomed in to have a better view, from which we can see the non-high income taxis’ distribution is relatively more even.
Fig 14.
The difference value distribution between high and non-high income taxis, the blue part represents the passengers that the high income taxis refuse to take, and the red part is the high income’s preferential passengers, and obviously, the areas of these two parts are the same.