Error when converting a pandas DataFrame to a GeoDataFrame

Question

I can't figure out how to convert a pandas DataFrame to a GeoDataFrame.

Here is an example of what my data looks like using df.head():

 crash_date. crash_time latitude longitude location 
0 2019年06月15日T00:00:00.000 14:57 40.8146250 -73.9203600 {'type': 'Point', 'coordinates': [-73.92036, 40.8146250]
1 2019年07月03日T00:00:00.000 0:50 40.8295970 -73.9022450 {'type': 'Point', 'coordinates': [-73.902245, 40.8295970] 
2 2019年06月24日T00:00:00.000 16:45 40.7054600 -73.7949000 {'type': 'Point', 'coordinates': [-73.7949, 40.7054600]
3 2019年06月16日T00:00:00.000 3:25 40.7128030 -73.9541700 {'type': 'Point', 'coordinates': [-73.95417, 40.7128030]

I tried converting it:

geometry = geometry=geopandas.points_from_xy(df.longitude, df.latitude)
df = df.drop(['longitude', 'latitude'], axis=1)
crs = {'init': 'epsg:4326'}
gdf = GeoDataFrame(df, crs=crs, geometry=geometry)

But I'm getting the following error:

---------------------------------------------------------------------------
TypeError Traceback (most recent call last)
<ipython-input-83-6090c239d222> in <module>
----> 1 geometry = geometry=gpd.points_from_xy(df.longitude, df.latitude)
 2 df = df.drop(['longitude', 'latitude'], axis=1)
 3 crs = {'init': 'epsg:4326'}
 4 gdf = GeoDataFrame(df, crs=crs, geometry=geometry)
 5 
~/opt/anaconda3/lib/python3.7/site-packages/geopandas/array.py in _points_from_xy(x, y, z)
 190 geom = [shapely.geometry.Point(i, j, k) for i, j, k in zip(x, y, z)]
 191 else:
--> 192 geom = [shapely.geometry.Point(i, j) for i, j in zip(x, y)]
 193 return geom
 194 
~/opt/anaconda3/lib/python3.7/site-packages/geopandas/array.py in <listcomp>(.0)
 190 geom = [shapely.geometry.Point(i, j, k) for i, j, k in zip(x, y, z)]
 191 else:
--> 192 geom = [shapely.geometry.Point(i, j) for i, j in zip(x, y)]
 193 return geom
 194 
~/opt/anaconda3/lib/python3.7/site-packages/shapely/geometry/point.py in __init__(self, *args)
 47 BaseGeometry.__init__(self)
 48 if len(args) > 0:
---> 49 self._set_coords(*args)
 50 
 51 # Coordinate getters and setters
~/opt/anaconda3/lib/python3.7/site-packages/shapely/geometry/point.py in _set_coords(self, *args)
 130 self._geom, self._ndim = geos_point_from_py(args[0])
 131 else:
--> 132 self._geom, self._ndim = geos_point_from_py(tuple(args))
 133 
 134 coords = property(BaseGeometry._get_coords, _set_coords)
~/opt/anaconda3/lib/python3.7/site-packages/shapely/geometry/point.py in geos_point_from_py(ob, update_geom, update_ndim)
 207 coords = ob
 208 n = len(coords)
--> 209 dx = c_double(coords[0])
 210 dy = c_double(coords[1])
 211 dz = None
TypeError: must be real number, not str

df.info() shows:

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 1000 entries, 0 to 999
Data columns (total 29 columns):
crash_date 1000 non-null object
crash_time 1000 non-null object
borough 620 non-null object
zip_code 620 non-null object
latitude 900 non-null object
longitude 900 non-null object
location 900 non-null object
on_street_name 782 non-null object
off_street_name 491 non-null object
number_of_persons_injured 1000 non-null object
number_of_persons_killed 1000 non-null object
number_of_pedestrians_injured 1000 non-null object
number_of_pedestrians_killed 1000 non-null object
number_of_cyclist_injured 1000 non-null object
number_of_cyclist_killed 1000 non-null object
number_of_motorist_injured 1000 non-null object
number_of_motorist_killed 1000 non-null object
contributing_factor_vehicle_1 994 non-null object
contributing_factor_vehicle_2 865 non-null object
collision_id 1000 non-null object
vehicle_type_code1 993 non-null object
vehicle_type_code2 812 non-null object
contributing_factor_vehicle_3 61 non-null object
contributing_factor_vehicle_4 23 non-null object
contributing_factor_vehicle_5 7 non-null object
vehicle_type_code_3 59 non-null object
vehicle_type_code_4 23 non-null object
vehicle_type_code_5 7 non-null object
cross_street_name 218 non-null object
dtypes: object(29)
memory usage: 226.7+ KB

Changed first line per recommendation

geometry = gpd.points_from_xy(df.longitude.values.astype('float32'), df.latitude.values.astype('float32'))
df = df.drop(['longitude', 'latitude'], axis=1)
crs = {'init': 'epsg:4326'}
gdf = gpd.GeoDataFrame(df, crs=crs, geometry=geometry)

Error message now shows:

AttributeError: 'DataFrame' object has no attribute 'longitude'

Answer 1

Your lon, lat cols are object types. Cast them to floats:

geometry = geopandas.points_from_xy(df.longitude.astype('float32'), df.latitude.astype('float32'))
#OR
geometry = geopandas.points_from_xy(df['longitude'].astype('float32'), df['latitude'].astype('float32'))

• jon: It's showing a different error now: AttributeError: 'DataFrame' object has no attribute 'longitude'
• PolyGeo: @jon Welcome to GIS SE! We're a little different from other sites; this isn't a discussion forum but a Q&A site. Please check out our short tour to learn about our focussed Q&A format. If it's a different error then that may be best asked as a new question.
• user2856: Make sure you're running that before you run df = df.drop(['longitude', 'latitude'], axis=1) and try df['longitude'].astype('float32'), df['latitude'].astype('float32')