Expand cities to European cities above 100k population

[mszell]

Inspired by #1
To do:

• Get European city names (not including the transcontinental countries for now: Russia, Kazakhstan, Azerbaijan, Georgia, and Turkey) from https://en.wikipedia.org/wiki/List_of_towns_and_cities_with_100,000_or_more_inhabitants, add them to a new file cities/european_100000pop.csv, like cities/european_capitals.csv. Start by taking the same nominatim_query column as the name column. Also add a new column population with the integer population, so can filter later by size, if needed.
• Double-check using the "Check city boundaries" section in scripts/build_recommends.ipynb that the nominatim_query really correspond to the intended city. If not, try to find the correct nominatim_query using nominatim.openstreetmap.org. If you can't find a working nominatim_query, note it down in Individual city issues for growbikenet #6. Also, while doing all this, also note down in Individual city issues for growbikenet #6 any cities with "weird" boundaries that OSM might provide. Let's define a boundary "weird" if it is not the same as in google maps (to be investigated/fixed at a later time).
• Double check boundaries with @mszell
• Once cities/european_100000pop.csv is ready, create a new csv called cities/european_capitalsand100000pop.csv by merging cities/european_100000pop.csv with cities/european_capitals.csv.
• Run the whole scripts/build_recommends.ipynb script with cityfilename = "european_capitalsand100000pop.csv". This will populate cities/rec_table.csv.

[mszell]

Please also add a column for the country, using alpha2.

[marianamirandapessoa]

The main data source of https://en.wikipedia.org/wiki/List_of_towns_and_cities_with_100,000_or_more_inhabitants is the UN's Demographic Yearbook 2022 > population > Table 8, so this Excel table was used to extract the names of the cities with more than 100,000 citizens.

The dataset was cleaned, and only the European cities were considered, resulting in 456 cities.

For each city, the name was used as the nominatim_query to inspect which polygons were extracted from OpenStreetMap.
Then, each extracted polygon was compared with the boundaries that appear on Google Maps.
For polygons with significant differences, the nominatim_query was adjusted to obtain the closest possible polygon match.

After the query adjustments, 360 polygons were accepted, and 96 weren't. Comments about the extracted polygons are present in european_100000pop_comments.xlsx.

The 96 "rejected" polygons are essentially cities where the municipality is larger than the urban fabric, and includes nearby parks, mountain ranges, or simply smaller towns that are less populated and less dense.

For each country, the number of "rejected" cities is:

Country	Number of "Rejected" Cities	All Cities in the Country
Bulgaria	2	5
Denmark	4	6
Greece	3	7
Italy	27	43
Portugal	3	6
Serbia	1	4
Spain	43	62
Switzerland	7	10
United Kingdom	6	29

[mszell]

I'm doing the last point myself: Run the whole scripts/build_recommends.ipynb script with cityfilename = "european_capitalsand100000pop.csv". This will populate cities/rec_table.csv