There is an R-Tree extension specifically for this purpose when working with fixed co-ordinate systems and multi-dimensional rectangles. <https://sqlite.org/rtree.html> There is a Geopoly extension specifically for this purpose when working with fixed co-ordinate systems and 2-dimensional polygons, which extends R-Tree by converting the N-sided polygons to 2-dimensional "bounding boxes" to quickly locate candidates. <https://sqlite.org/geopoly.html> As for Python and Panda dataframe, have reference to the documentation for those whacks of code, neither of which are SQLite3.