The City column looks good and contains exactly one value. The Geolocation column, on the other hand, contains four variables: latitude, latitude direction, longitude, and longitude direction. Let's split the Geolocation column into four separate columns:
Because the original data type for the Geolocation was an object, all the new columns are also objects. Let's change latitude and longitude into floats:
>>> geolocations = geolocations.astype({'latitude':'float', 'longitude':'float'}) >>> geolocations.dtypes latitude float64
latitude direction object
longitude float64
longitude direction object
dtype: object
Concatenate these new columns with the City column from the original: