Checkout https://aakashjhawar.github.io/commercial-centers-using-POI/
Identify commercial centers using Points of Interest (POI) data by clustering these points into commercial centers/markets.
How to use
git clone git@github.com:aakashjhawar/commercial-centers-using-POI.git
cd commercial-centers-using-POI
pip3 install -r requirements.txt
jupyter notebook
Start the jupyter server. The following should be executed in order.
- gather-data.ipynb
- DBSCAN-remove-noise.ipynb
- create-clusters.ipynb
It uses Overpy (Python wrapper for Overpass) to gather all the nodes of a particular location and stores the data into a csv. It also cleans the dataset based on amenities.
It removes the noise/outliers present in the dataset. It also gives the number of clusters that can be formed from the dataset.
It uses KMeans Clustering to create clusters and plot them on Google Map using gmplot library.
Check out the Github repo page