QARTA: An ML-based System for Accurate Map Services

This repository contains the code for experiments of QARTA components along with the code for the query calibration module.

To start, follow these steps:

create a conda envrionment based on 'environment file' in this repo as follows:

conda env create -f environment.yml

Next, prepare your trips dataset to be as follows:

TripStartTime,PickupLon,PickupLat,DropLon,DropLat,GT_Distance,GT_Duration,OsrmDistance,OsrmDuration

The date should be fomrated as (month-day-year hour:minutes), python:('%m-%d-%y %H:%M'). A sample file for NYC trips is given in the data folder.

The last two column OsrmDistance and OsrmDuration can be obtained by querying OSRM. A dockerized version is abvailable in thie repo.

Next, run data_prep.py to prepare the spatial zoning and obtain OSRM results.

Make sure to update the file names for your data, and the zones files. Also make sure you configure the dockerfile to download the map of the state/country that you are working on.

Once prepared, run the query_calibration.py.