Open brezniczky opened 5 years ago
Mentioned in the Kaggle script, the github version is a #TODO.
By the way, this could be adjusted for by splitting the data set into the two subsets (Prussia etc.), then analysing the parts individually.
Should comment about this on the Kaggle notebook.
This one:
https://www.reddit.com/r/europe/comments/7fwq02/polish_election_results_overlaid_on_a_map_of_the/
"Polish election results overlaid on a map of the former German Empire"
Not only I am not the best person to comment on their results, this underlines that there can be options to find two poles of political preference (as mentioned by others). Maybe something (such as ward sizes) also correlates with the distribution of votes on a geographic level.