How to make necessary changes to Geodataframe to export as geopackage or shapefile

Question

I am trying to export a geodatabase I created via the OMSNX package. ()

When I attempt to convert the file from a geodatabase to shapefile I get an error "ValueError: Invalid field type <class 'list'>". This is because the geodatabase has some formats that are not compatible with shapefiles as mentioned here under the spatial data section.

How do I convert the geodataframe to allow for an export as a shapefile? My code and attached error are below.

import geopandas as gpd
import osmnx as ox
import fiona as fi
print(fi.supported_drivers)
%matplotlib inline
ox.__version__

place = "Dickinson, Texas"
tags = {"building": True}
gdf = ox.geometries_from_place(place, tags)
gdf.shape



fig, ax = ox.plot_footprints(gdf, figsize=(8, 8))
gdf.shape
gdf.to_file('testgalvy.gpkg',driver='GPKG')

Answer 1

Whenever I get "ValueError: Invalid field type <class 'list'>". it is when I am trying to convert a geodataframe which has a string/object field with lists in it to shapefile. Do you have lists in a column?

You can convert them to strings by doing:

import pandas as pd
df = pd.DataFrame({'a':[1,2,3],'b':[['a','list','in','the','column'],'b','c']}) #Create some test data


   a                           b
0  1  [a, list, in, the, column]
1  2                           b
2  3                           c

df['b'] = df['b'].apply(lambda x: ' '.join(x)) #Join the list into a string

   a                     b
0  1  a list in the column
1  2                     b
2  3                     c

If you have alot of data you can check all columns like this:

for col in df.columns:
    if any(isinstance(val, list) for val in df[col]):
        print('Column: {0}, has a list in it'.format(col))
Column: b, has a list in it