Error converting geometry column from string type - GeoPandas

Question

I have a pandas DataFrame with geometry column with Polygons. I am trying to convert it to a geopandas dataframe and need to change the column type of geometry from string.

I get this error when attempting to apply Polygon from shapely.

import geopandas as gpd 
from shapely.geometry import Polygon

df_geo['geometry'][0]

POLYGON ((-83.230504 42.242044, -83.228256 42.242185, -83.222044 42.242297, -83.221459 42.242307, -83.21916399999999 42.242432, -83.21682 42.242508, -83.21650799999999 42.23718, -83.216279 42.232113, -83.216138 42.227846, -83.219548 42.227773, -83.229843 42.227545, -83.229902 42.228622, -83.230504 42.242044))
                     

# convert geometry to Polygon

df_geo['geometry'] = df_geo['geometry'].apply(lambda x: Polygon(eval(x)))

df_geo



Traceback (most recent call last):

  File "/Applications/Anaconda/anaconda3/lib/python3.9/site-packages/IPython/core/interactiveshell.py", line 3444, in run_code
    exec(code_obj, self.user_global_ns, self.user_ns)

  File "/var/folders/d0/gnksqzwn2fn46fjgrkp6045c0000gn/T/ipykernel_23632/494222449.py", line 3, in <module>
    df_geo['geometry'] = df_geo['geometry'].apply(lambda x: Polygon(eval(x)))

  File "/Applications/Anaconda/anaconda3/lib/python3.9/site-packages/pandas/core/series.py", line 4357, in apply
    return SeriesApply(self, func, convert_dtype, args, kwargs).apply()

  File "/Applications/Anaconda/anaconda3/lib/python3.9/site-packages/pandas/core/apply.py", line 1043, in apply
    return self.apply_standard()

  File "/Applications/Anaconda/anaconda3/lib/python3.9/site-packages/pandas/core/apply.py", line 1098, in apply_standard
    mapped = lib.map_infer(

  File "pandas/_libs/lib.pyx", line 2859, in pandas._libs.lib.map_infer

  File "/var/folders/d0/gnksqzwn2fn46fjgrkp6045c0000gn/T/ipykernel_23632/494222449.py", line 3, in <lambda>
    df_geo['geometry'] = df_geo['geometry'].apply(lambda x: Polygon(eval(x)))

  File "<string>", line 1
    POLYGON ((-83.230504 42.242044, -83.228256 42.242185, -83.222044 42.242297, -83.221459 42.242307, -83.21916399999999 42.242432, -83.21682 42.242508, -83.21650799999999 42.23718, -83.216279 42.232113, -83.216138 42.227846, -83.219548 42.227773, -83.229843 42.227545, -83.229902 42.228622, -83.230504 42.242044))
                         ^
SyntaxError: invalid syntax

Answer 1

Your polygon string is in wkt format. You can use shapely.wkt.loads:

import pandas as pd
import geopandas as gpd
import shapely

data = [['A', 'POLYGON ((-83.230504 42.242044, -83.228256 42.242185, -83.222044 42.242297, -83.221459 42.242307, -83.21916399999999 42.242432, -83.21682 42.242508, -83.21650799999999 42.23718, -83.216279 42.232113, -83.216138 42.227846, -83.219548 42.227773, -83.229843 42.227545, -83.229902 42.228622, -83.230504 42.242044))']]
df = pd.DataFrame.from_records(data=data, columns=['name','g'])

df['geometry'] = df['g'].apply(lambda x: shapely.wkt.loads(x))
gdf = gpd.GeoDataFrame(data=df, geometry=df['geometry'], crs=4329)

If you have invalid wkt's you can find and drop them like this, prior to the code above:

import pandas as pd
import geopandas as gpd
import shapely
from shapely.errors import WKTReadingError

data = [['A', 'POLYGON ((-83.ZXCZXZCXZCZX30504 42.242044, -83.228256 42.242185, -83.222044 42.242297, -83.221459 42.242307, -83.21916399999999 42.242432, -83.21682 42.242508, -83.21650799999999 42.23718, -83.216279 42.232113, -83.216138 42.227846, -83.219548 42.227773, -83.229843 42.227545, -83.229902 42.228622, -83.230504 42.242044))'],
        ['B', 'POLYGON ((-83.230504 42.242044, -83.228256 42.242185, -83.222044 42.242297, -83.221459 42.242307, -83.21916399999999 42.242432, -83.21682 42.242508, -83.21650799999999 42.23718, -83.216279 42.232113, -83.216138 42.227846, -83.219548 42.227773, -83.229843 42.227545, -83.229902 42.228622, -83.230504 42.242044))']]
df = pd.DataFrame.from_records(data=data, columns=['name','g'])

invalid = []

for ix in df.index:
    try:
        shapely.wkt.loads(df.iloc[ix].g) 
    except WKTReadingError: 
        invalid.append(ix) #If it is invalid, append its index to invalid list.
        
df = df.drop(invalid) #And delete all invalid rows