NYC Taxi Ride Sharing

The project aims to develop a real-time ride sharing system to develop & study algorithms to optimally share rides among users.

Members

1. Benito Alvares
2. Harish Ventaktaraman
3. Karan Venkatesh Davanam
4. Prajwal Kishor Kammardi

Components needed

1. PostgreSQL installed
2. PostGIS extension installed in the relevant schema
3. Python 3.6
4. pip or conda virtual environments
5. Components specified in the requirements.txt or conda_requirements.txt

PostgreSQL installation instructions

1. Have a Role called nycrideshare with password nycrideshare created in the environment. (You can also change the python file to specify the ID/Password of you environments)
2. Create a new database

CREATE DATABASE nyc_taxi
    WITH
    OWNER = nycrideshare
    ENCODING = 'UTF8'
    LC_COLLATE = 'English_United States.1252'
    LC_CTYPE = 'English_United States.1252'
    TABLESPACE = pg_default
    CONNECTION LIMIT = -1;

3. Create schema called as nyc_taxi_schema and public (if not present already)

CREATE SCHEMA nyc_taxi_schema
    AUTHORIZATION nycrideshare;

CREATE SCHEMA public
    AUTHORIZATION nycrideshare;

4. Install PostGIS. After installation add the extension using the query below. Check if the tables and functions are created under the public schema before and after running the below query.

CREATE EXTENSION postgis
    SCHEMA public
    VERSION "3.1.1";

5. Run the final_project.ipynb to create the tables needed for the project and populate them.

6. Create a Stored function in PostgreSQL.

CREATE OR REPLACE FUNCTION nyc_taxi_schema.get_cust_between_timestamps_lgd(IN timevalue text DEFAULT ''::text,IN timeinterval text DEFAULT  '5 MINUTES'::text)
    RETURNS TABLE(id integer, tpep_pickup_datetime timestamp without time zone, tpep_dropoff_datetime timestamp without time zone, passenger_count integer, "PULocationID" integer, "DOLocationID" integer)
    LANGUAGE 'plpgsql'
    VOLATILE
    PARALLEL SAFE
    COST 100    ROWS 1000

AS $BODY$
BEGIN
RETURN QUERY
	SELECT
			rd.id,
			rd.tpep_pickup_datetime,
			rd.tpep_dropoff_datetime,
			rd.passenger_count,
			rd."PULocationID",
			rd."DOLocationID"
	FROM
		nyc_taxi_schema.ride_details as rd

	WHERE
		(
			rd."PULocationID" = 138
			OR
			rd."DOLocationID" = 138
		)
		AND
		(
			rd.tpep_pickup_datetime BETWEEN timeValue::timestamp
			AND
			timeValue::timestamp + timeInterval::INTERVAL
		)
		AND
		(
			rd.passenger_count < 3
		)
		ORDER BY rd.tpep_pickup_datetime ASC;
END
$BODY$;

7. Build Index in Postgres for the pickup timestamp.

CREATE INDEX pickup_time_index
    ON nyc_taxi_schema.ride_details USING brin
    (tpep_pickup_datetime)
;

Geopandas installation instructions for Windows

1. These steps assume that you have installed wheel (pip install *.whl on cmd).

2. Go to Unofficial Windows Binaries for Python Extension Packages.

3. Download on a specific folder the following binaries: GDAL, Pyproj, Fiona, Shapely and Geopandas matching the version of Python, and whether the 32-bit or 64-bit OS is installed on your laptop. (E.g. for Python v3.7x (64-bit), GDAL package should be GDAL‑3.1.2‑cp37‑cp37m‑win_amd64.whl.)

4. Use Command Prompt and go to the folder where you have downloaded the binaries

5. Important: The following order of installation using pip install is necessary. Be careful with the filename. It should work if the filename is correct: (Tip: Type “pip install” followed by a space and type the first two letters of the binary and press Tab. (e.g. pip install gd(press Tab))

• pip install .\GDAL-3.1.1-xxx.whl

• pip install .\pyproj-2.6.1.post1-xxx.whl

• pip install .\Fiona-1.8.13-xxx.whl

• pip install .\Shapely-1.7.0-xxx.whl

• pip install .\geopandas-0.8.0-py3-none-any

  replace xxx with your python version.

To run the project.

1. Make sure that the OSRM backend is up and running. Instructions are in this link
2. Run the final_project.ipynb file to populate the database
3. Run nyc_ride_simulations to run the simulations interactively. Note that the features are limited and it is just for demo purposes.
4. To run a proper simulation, go to run folder.
   1. use python run.py "2019-01-01 00:00:00" to run simulations.
   2. use python viz.py to see the visualizations.