Geopandas: GeoDataFrame .loc method returns pandas Series instead of GeoSeries?

[dzanaga]

If I select a row with .loc from a GeoDataFrame, I get out a Series instead of a GeoSeries. Is this an intended behavior?

[Rabscuttler]

You will only get a GeoSeries if all values in the selected row or column are shapely geometric objects, from the docs here: http://geopandas.org/data_structures.html

So I think it is intended.

The code snippet below shows this behaviour:

import pandas as pd
import geopandas as gpd

df = pd.DataFrame( {'City': ['Buenos Aires', 'Brasilia', 'Santiago', 'Bogota', 'Caracas'], 'Country': ['Argentina', 'Brazil', 'Chile', 'Colombia', 'Venezuela'], 'Latitude': [-34.58, -15.78, -33.45, 4.60, 10.48], 'Longitude': [-58.66, -47.91, -70.66, -74.08, -66.86]})

gdf = gpd.GeoDataFrame( df, geometry=gpd.points_from_xy(df.Longitude, df.Latitude))

# row is pd.Series
print(type(gdf.loc[0]))

# column is geopandas.Series if all values are Points, Lines or Polygons (or Multi-*)
print(type(gdf.loc[:,'City']))
print(type(gdf.loc[:,'geometry']))

[martinfleis]

As @Rabscuttler nicely explained, GeoSeries is "a vector where each entry in the vector is a set of shapes corresponding to one observation", so this is indeed intended behaviour. I will close this issue now, if we want to discuss it more we can always reopen.

[raphaelquast]

... just in case someone stumbles upon this issue while searching for an efficient way to get a single-row geodataframe from an existing geodataframe (as i just did)... this does the trick:

df1 = df.iloc[1]    
>>> a pandas Series of the first row

df1 = df.iloc[[1]]
>>> a GeoDataframe with only the first row