Identifying commercial centers using Points of Interest (POI) OSM data for New Delhi using DBSCAN (Density-Based Spatial Clustering of Applications with Noise) algorithm.
The following map highlights the results (all the commercial centers/markets) obtained after analysis marked on New Delhi's map.
The following scatter plot provides an overview of the markets(clusters) identified during analysis.
x-axis - Longitude, y-axis - Latitude
-
Setup a conda environment or virtualenv.
-
Install requirements.txt.
Steps:-
$ pip install virtualenv$ virtualenv myenv$ myenv\Scripts\activate$ pip install -r requirements.txt
- clustered.csv - Contains cluster labels for all the commercial nodes.
- markets.csv - Contains centre points for all the clusters/markets identified in New Delhi region.
- data_manip.ipynb - Collection, cleaning and visualization of data obtained using overpass API, pandas and numpy.
- clustering.ipynb - Algorithm for clustering. Analysis and visualization of results using Scatter Plots, Bar Plots, gmplot and Reverse Geocoding.
- map_delhi.html - Market coordinates plotted on the map of New Delhi.
- map_delhi_{1,2,3,4}.html - Markets identified for different values of epsilon during analysis.
- market_coor_shape_files.zip - zip file containing shape file for coordinates of identified markets.
- requirements.txt - Dependencies required for running the scripts.
- All other json and pickle files contain data obtained after different stages during the analysis.