https://twitter.com/allixender/status/1323710840273543175
It is a dataset provided by the Information Technology and Development Center of the Ministry of the Interior of Estonia (SMIT). SMIT is the largest IT institution in the country, which creates and manages information systems necessary for saving lives and ensuring internal security. The data contains anonymised records of crimes, felonies and other offenses against the law. The data is binned into squares of either 500mx500m or 1000mx1000m.
https://opendata.riik.ee/et/dataset/avaliku-korra-vastased-ja-avalikus-kohas-toime-pandud-syyteod
https://opendata.smit.ee/ppa/csv/avalik_3.csv
In addition I took the Administrative-and-Settlement-Divisions dataset from the Estonian Landboard and only took the Tallinn area.
https://geoportaal.maaamet.ee/eng/Spatial-Data/Administrative-and-Settlement-Division-p312.html
https://geoportaal.maaamet.ee/docs/haldus_asustus/asustusyksus_shp.zip?t=20201101020927
https://geoportaal.maaamet.ee/index.php?lang_id=2&page_id=663
Most challenging was the pre-processing of the crime dataset. As the polygons (squares) are a bit too small to look great on Estonian national level, I decided to overlay it on the Tallinn city and sub-urban area. In addition I classified the data with PySAL’s mapclassify and used Contextily to provide a static basemap.
Again the main work is done with Pandas/GeoPandas and matplotlib.
The Jupyter notebook (view here) is provided in my GitHub repo.