ValueError: Cannot render objects with any missing geometries when using a geopandas dataframe

Question

I am trying to add a choropleth layer on a folium map using a geopandas dataframe that I built by merging a shapefile containing all the regions in Brazil (named "map_1"), with a regular pandas dataframe (named "amazon_state"). The merging of the 2 dataframes gives me "map_2", which after cleaning (removal of some rows) I referred to as "map_3".

'''

#importing shapefile map_1, which contains all regions in Brazil

map_1= gpd.read_file("/Users/alexandertankou/Desktop/python/bra_adm1/BRA_adm1.shp")

# creating amazon_state

amazon_state= amazon_data.groupby("state", as_index=False).sum().drop(columns=["year"])

# ensure the naming of regions e in map_1 and amazon_state is the same

map_1.NAME_1= amazon_state.state

#map_2: merging map_1 with amazon_state
map_2= pd.merge (left=map_1, right= amazon_state, left_on="NAME_1", right_on="state", how= "left")

#dropping none useful columns
map_2= map_2.drop(columns=["NAME_0",'HASC_1',"ID_1","ISO","CCN_1","CCA_1","ID_0","TYPE_1","ENGTYPE_1", "NL_NAME_1","VARNAME_1"])

#ploting map_2
map_2.plot(column="number", cmap="YlOrRd",legend=True, figsize= (13,10))

#setting the folium map

m=folium.Map(location= [-22.919882,-43.604392], zoom_start=10)

#making geomery type hashable in python
map_2['geometry'] = map_2['geometry'].apply(lambda x: str(x))

#cleaning up map_2 data
map_3= map_2.iloc[0:23,:]

#add the choropleth layer on folium map

m.choropleth(geo_data= map_3,name="geometry", data= map_3,key_on="feature.properties.NAME_1",columns=["geometry","number"],fill_color='YlGn')
folium.LayerControl().add_to(m)

'''

But I keep getting the ValueError: Cannot render objects with any missing geometries. Having used isnan and is_empty methods, I know for sure there are no missing values in map_3 (see the data below) so not sure what I am doing wrong:

              NAME_1                                           geometry  \
0               Acre  POLYGON ((-73.33251190185541 -7.32487916946411...   
1            Alagoas  MULTIPOLYGON (((-35.90152740478516 -9.86180496...   
2              Amapa  MULTIPOLYGON (((-50.02402877807612 0.859862029...   
3           Amazonas  POLYGON ((-67.32623291015625 2.029680967331046...   
4              Bahia  MULTIPOLYGON (((-38.69708251953125 -17.9790287...   
5              Ceara  MULTIPOLYGON (((-38.47541809082026 -3.70097303...   
6   Distrito Federal  POLYGON ((-48.03603363037109 -15.5002202987670...   
7     Espirito Santo  MULTIPOLYGON (((-40.88402938842768 -21.1612491...   
8              Goias  POLYGON ((-50.15817260742188 -12.4123792648315...   
9           Maranhao  MULTIPOLYGON (((-42.12374877929688 -2.80069398...   
10       Mato Grosso  POLYGON ((-56.1036376953125 -17.17354011535639...   
11      Minas Gerais  POLYGON ((-57.6052360534668 -8.662846565246525...   
12           Paraiba  POLYGON ((-44.20977783203119 -14.2366542816162...   
13              Par·  MULTIPOLYGON (((-46.43458175659174 -1.01708304...   
14        Pernambuco  MULTIPOLYGON (((-42.87873840332026 -9.29837322...   
15              Piau  MULTIPOLYGON (((-48.63069534301758 -25.8679161...   
16               Rio  MULTIPOLYGON (((-35.13597106933594 -8.83791732...   
17          Rondonia  POLYGON ((-41.81680679321283 -2.74375009536743...   
18           Roraima  MULTIPOLYGON (((-44.67124938964838 -23.3545837...   
19    Santa Catarina  MULTIPOLYGON (((-35.10902786254883 -6.19347190...   
20         Sao Paulo  MULTIPOLYGON (((-52.07069396972656 -32.0284729...   
21           Sergipe  POLYGON ((-63.53470230102539 -7.97433900833129...   
22         Tocantins  POLYGON ((-60.16886138916004 5.226301193237362...   

               state     number  
0               Acre  18464.030  
1            Alagoas   4644.000  
2              Amapa  21831.576  
3           Amazonas  30650.129  
4              Bahia  44746.226  
5              Ceara  30428.063  
6   Distrito Federal   3561.000  
7     Espirito Santo   6546.000  
8              Goias  37695.520  
9           Maranhao  25129.131  
10       Mato Grosso  96246.028  
11      Minas Gerais  37475.258  
12           Paraiba  52435.918  
13              Par·  24512.144  
14        Pernambuco  24498.000  
15              Piau  37803.747  
16               Rio  45160.865  
17          Rondonia  20285.429  
18           Roraima  24385.074  
19    Santa Catarina  24359.852  
20         Sao Paulo  51121.198  
21           Sergipe   3237.000  
22         Tocantins  33707.885

Answer 1

@Thomas is right: your problem is your map_2 and map_3 variables corresponds to <class 'pandas.core.frame.DataFrame'>, but you need <class 'geopandas.geodataframe.GeoDataFrame'>.

The reason is here:

map_2 = pd.merge(left=map_1, right=amazon_state, left_on="NAME_1",
                 right_on="state", how= "left")

When you call pandas.merge() method and your arguments are GeoDataFrame and DataFrame the result will be always DataFrame.

If you want to get GeoDataFrame after merging use pandas.DataFrame.merge() method instead. Make sure you call it from the GeoDataFrame, not DataFrame:

map_2 = map_1.merge(right=amazon_state, left_on="NAME_1",
                    right_on="state", how="left")

You could also check relating documentation for geopandas.

Answer 2

i had the same problem. map_3 is a geodataframe ? if not, you have to convert before :

gdf = gpd.GeoDataFrame(map_3 , geometry = map_3.geometry)
gdf.crs = {'init' :'epsg:4326'}