Import geonames populated places into a dataset http://download.geonames.org/export/dump/

geonames_populated_places.py

#!/usr/bin/python3

import pandas as pd
import numpy as np

dtypes_dict = {
    0: np.dtype(int), # geonameid
    1: np.dtype(str),  # name
    2: np.dtype(str),  # asciiname
    3: np.dtype(str),  # alternatenames
    4: np.dtype(float), # latitude
    5: np.dtype(float), # longitude
    6: np.dtype(str), # feature class
    7: np.dtype(str), # feature code
    8: np.dtype(str), # country code
    9: np.dtype(str), # cc2
    10: np.dtype(str), # admin1 code
    11: np.dtype(str), # admin2 code
    12: np.dtype(str), # admin3 code
    13: np.dtype(str), # admin4 code
    14: np.dtype(int), # population
    15: np.dtype(float), # elevation
    16: np.dtype(str), # dem (digital elevation model)
    17: np.dtype(str), # timezone
    18: np.dtype(str) # modification date yyyy-MM-dd
}

df_chunk = pd.read_csv('allCountries.txt', sep="\t", header = None, dtype=dtypes_dict, chunksize=1000000)
chunk_list = []
for chunk in df_chunk:
    chunk_list.append(chunk)
geonames = pd.concat(chunk_list)

places = geonames[geonames[6] == 'P']

dataset = places.filter([4,5,1,10,11,8])
dataset.to_csv('geonames.csv', header=['lat','lon','name','admin1','admin2','cc'], sep=',',index=False)