Parkings and weather data correlation and forecasting

This project is for analysing and predicting parking usage rates in the City of Turku. Weather data is used as an extra feature to analyze correlation and improve predictions.

The project has been bootstrapped with uv https://docs.astral.sh/uv/.

How to

Data

Get necessary csv data and place it in root's data folder. All csv files in data folder are gitignored. Following are base data files:

• parkings_small.csv
• artukainen_rain.csv
• artukainen_temperature.csv

File parkings_small.csv contains city of Turku paid parkings data from 2024-05-07 to 2025-02-06. The necessary data columns are time_start, time_end, zone_id and parking_area_id.

• time_start and time_end tells when a single parking has occurred.
• zone_id tells which parking zone the parking was in.
• parking_area_id tells which parking area the parking was in.

Files artukainen_rain and artukainen_temperature contains hourly timestamped rain and temperature values.

uv

To use uv, install it first (https://docs.astral.sh/uv/getting-started/installation/#installation-methods) e.g.:

curl -LsSf https://astral.sh/uv/install.sh | sh

Test everything works:

uv run main.py

Update used packages by running:

uv sync

Activate venv:

source .venv/bin/activate