Python package implementing fetching, computation and visualisation of data related to Warsaw public bus transit. Implemented as a final project for NYPD classes at University of Warsaw, February 2020.
Positions are fetched by repetitive querying bus positions API. Script queries the API every 15 seconds, aggregates the results and removes the duplicates. Results are then saved in the csv file.
# To fetch positions from given period of time
python main.py positions --api_key XYZ --file positions.csv --start "2021-02-10 16:00" --end "2021-02-10 18:00"
Example data
| Lines | Brigade | Lon | Lat | VehicleNumber | Time |
|---|---|---|---|---|---|
| 130 | 1 | 21.0841126 | 52.1676473 | 1001 | 2021-02-02 16:29:50 |
| 213 | 4 | 21.1004743 | 52.226027 | 1002 | 2021-02-02 16:29:51 |
| 213 | 2 | 21.214459 | 52.1616376 | 1003 | 2021-02-02 16:29:52 |
Schedules are fetched by first querying the API for all bus stops, then querying for lines visiting each stop, then for the detailed schedule of those lines.
Note, that the schedule may differ on the day of the script execution - it differs between workdays and weekends.
# To fetch schedules
python main.py schedules --api_key XYZ --file schedules.csv
Example data
| Lines | Brigade | Lon | Lat | BusStopName | Time |
|---|---|---|---|---|---|
| 123 | 4 | 21.044827 | 52.248455 | Kijowska | 04:49:00 |
| 123 | 6 | 21.044827 | 52.248455 | Kijowska | 05:19:00 |
| 123 | 357 | 21.044827 | 52.248455 | Kijowska | 05:49:00 |
Speeds are calculated based on the previously fetched positions for a given time period. First position of a given bus is NaN, each consecutive is based on the distance traveled between data points of this bus.
Data is then visualized in the form of splatter map.
# To calculate speeds
python main.py speeds --file speeds.csv --positions positions.csv --start "2021-02-10 16:00" --end "2021-02-10 18:00"
# To just visualize previously calculated speeds
python main.py speeds --file speeds.csv --dry
Example data
| Lines | Brigade | Lon | Lat | VehicleNumber | Time | Distance | Speed |
|---|---|---|---|---|---|---|---|
| 150 | 010 | 20.925739 | 52.310017 | 733 | 2021-02-02 17:00:08 | NaN | NaN |
| 150 | 010 | 20.927332 | 52.308876 | 733 | 2021-02-02 17:00:27 | 0.1671 | 31.661 |
| 150 | 010 | 20.927723 | 52.308624 | 733 | 2021-02-02 17:00:34 | 0.0386 | 19.901 |
Example visualization
On 2021-02-02 between 17:00:00 and 18:00:00 555/1544 of the buses exceeded
60 km/h at least once. Map suggests that it usually occurs on high-speed roads
and outside city centre.
Delays are calculated based on both the previously fetched positions and schedules for a given time period. Arrival is based on the first occurrence of an adequate bus (determined by bus line and brigade) in the radius of 250 meters from the bus stop, considering only the arrivals at most 5 minutes before scheduled time.
# To calculate delays
python main.py delays --file delays.csv --positions positions.csv --schedules schedules.csv --start "2021-02-10 16:00" --end "2021-02-10 18:00"
# To just visualize previously calculated speeds
python main.py delays --file delays.csv --dry
Example data
| Lines | Brigade | Lon | Lat | BusStopName | Time | Delay |
|---|---|---|---|---|---|---|
| 109 | 3 | 20.980452 | 52.229405 | rondo Daszyńskiego | 2021-02-02 17:37:00 | 2.0 |
| 109 | 3 | 20.970592 | 52.228602 | Szpital Wolski | 2021-02-02 17:35:00 | 2.0 |
| 109 | 3 | 20.958139 | 52.227852 | PKP Wola (Kasprzaka) | 2021-02-02 17:33:00 | 1.0 |
Example visualization
On 2021-02-02 between 17:00:00 and 18:00:00 mean delay is just 1.34 minutes.
Tests can be run using pytest command.
$ pytest --cov=warsawbus
====================== test session starts =====================================
platform darwin -- Python 3.6.8, pytest-6.2.1, py-1.10.0, pluggy-0.13.1
rootdir: /Users/abukala/IdeaProjects/bus
plugins: requests-mock-1.8.0, cov-2.11.1
collected 15 items
warsawbus/fetch/test_fetchers.py ............ [ 80%]
warsawbus/statistics/test_calculators.py ... [100%]
---------- coverage: platform darwin, python 3.6.8-final-0 -----------
Name Stmts Miss Cover
--------------------------------------------------------------
warsawbus/__init__.py 4 0 100%
warsawbus/fetch/__init__.py 4 0 100%
warsawbus/fetch/fetcher.py 25 10 60%
warsawbus/fetch/position_fetcher.py 20 0 100%
warsawbus/fetch/schedule_fetcher.py 44 1 98%
warsawbus/fetch/test_fetchers.py 81 0 100%
warsawbus/statistics/__init__.py 4 0 100%
warsawbus/statistics/calculator.py 10 1 90%
warsawbus/statistics/delay_calculator.py 43 0 100%
warsawbus/statistics/speed_calculator.py 25 0 100%
warsawbus/statistics/test_calculators.py 40 0 100%
warsawbus/visualize/__init__.py 2 0 100%
warsawbus/visualize/plotter.py 40 27 32%
--------------------------------------------------------------
TOTAL 342 39 89%
======================== 15 passed in 4.20s ====================================
To build a source distribution run
python setup.py sdist
To install package run
pip install $PATH
geopy
pandas
numpy
plotly
pytest