5 Best Ways to Convert Pandas DataFrame to GeoPandas

by

💡 Problem Formulation: Geospatial analysis is an integral part of modern data analysis. With the prolific use of Python for data science, it’s often necessary to convert a standard pandas DataFrame containing geographic coordinates into a GeoDataFrame for spatial analysis using GeoPandas. This article illustrates how to perform this conversion effectively. The input is a pandas DataFrame with columns representing latitude and longitude, and the desired output is a GeoDataFrame ready for geospatial computations.

Method 1: Using the GeoDataFrame() Constructor

This method entails creating a GeoDataFrame object directly by passing the existing pandas DataFrame to the GeoDataFrame constructor along with a geometry series made by zipping the longitude and latitude columns. This is the most straightforward approach when you have simple point data.

Here’s an example:

import pandas as pd
import geopandas as gpd
from shapely.geometry import Point

# Sample DataFrame with longitude and latitude
df = pd.DataFrame({
    'City': ['New York', 'Los Angeles', 'Chicago'],
    'Longitude': [-74.0060, -118.2437, -87.6298],
    'Latitude': [40.7128, 34.0522, 41.8781]
})

# Converting DataFrame to GeoDataFrame
gdf = gpd.GeoDataFrame(df, geometry=[Point(xy) for xy in zip(df['Longitude'], df['Latitude'])])
print(gdf)

Output:

          City  Longitude  Latitude                    geometry
0     New York    -74.0060   40.7128  POINT (-74.006 40.7128)
1  Los Angeles   -118.2437   34.0522 POINT (-118.2437 34.0522)
2      Chicago    -87.6298   41.8781  POINT (-87.6298 41.8781)

This code snippet first imports the required libraries: pandas, GeoPandas, and Shapely for geometrical operations. A pandas DataFrame containing cities and their respective longitudes and latitudes is created, which is then converted into a GeoDataFrame by pairing the longitude and latitude columns to create Point objects, which are used as the geometry argument in the GeoDataFrame constructor.

Method 2: Using the set_geometry() Function

The set_geometry() function in GeoPandas allows for the conversion of a pandas DataFrame by setting the geometry column explicitly. This method is useful when the DataFrame already contains a geometry column or when you need to change the active geometry of an existing GeoDataFrame.

Here’s an example:

import pandas as pd
import geopandas as gpd
from shapely.geometry import Point

# Sample DataFrame with longitude and latitude
df = pd.DataFrame({
    'City': ['New York', 'Los Angeles', 'Chicago'],
    'Coords': [Point(-74.0060, 40.7128), Point(-118.2437, 34.0522), Point(-87.6298, 41.8781)]
})

# Converting DataFrame to GeoDataFrame
gdf = gpd.GeoDataFrame(df).set_geometry('Coords')
print(gdf)

Output:

          City                      Coords
0     New York  POINT (-74.006 40.7128)
1  Los Angeles POINT (-118.2437 34.0522)
2      Chicago  POINT (-87.6298 41.8781)

After importing the necessary libraries, a pandas DataFrame is created with a ‘Coords’ column containing Point objects representing city locations. Using the set_geometry() function on the resulting GeoDataFrame, ‘Coords’ is established as the geometry column. The GeoDataFrame is now spatially enabled and ready for further geospatial analysis.

Method 3: Integrating with apply() and a Custom Function

This method uses pandas’ apply() function together with a custom function to apply the Point constructor to each row of the DataFrame individually. It’s particularly useful when custom manipulations on the coordinate columns are needed before creating Point objects.

Here’s an example:

import pandas as pd
import geopandas as gpd
from shapely.geometry import Point

# Sample DataFrame with coordinates
df = pd.DataFrame({
    'City': ['New York', 'Los Angeles', 'Chicago'],
    'Longitude': [-74.0060, -118.2437, -87.6298],
    'Latitude': [40.7128, 34.0522, 41.8781]
})

# Custom function to create Point geometry
def create_point(row):
    return Point(row['Longitude'], row['Latitude'])

# Applying custom function to create geometry column
df['geometry'] = df.apply(create_point, axis=1)

# Converting DataFrame to GeoDataFrame
gdf = gpd.GeoDataFrame(df, geometry='geometry')
print(gdf)

Output:

          City  Longitude  Latitude                    geometry
0     New York    -74.0060   40.7128  POINT (-74.006 40.7128)
1  Los Angeles   -118.2437   34.0522 POINT (-118.2437 34.0522)
2      Chicago    -87.6298   41.8781  POINT (-87.6298 41.8781)

The code example utilizes the apply method of pandas to iterate over each row of the DataFrame and apply the custom create_point() function, which takes a row as input and returns a Point geometry object created from the ‘Longitude’ and ‘Latitude’ columns. This approach is especially versatile for more complex row-wise operations before converting to a GeoDataFrame.

Method 4: Applying a Lambda Function

Lambda functions in Python offer a quick, inline alternative to define a function. When combined with the apply() feature of pandas, this allows for streamlined and readable code for creating Point geometries before conversion to a GeoDataFrame.

Here’s an example:

import pandas as pd
import geopandas as gpd
from shapely.geometry import Point

# Sample DataFrame with longitude and latitude
df = pd.DataFrame({
    'City': ['New York', 'Los Angeles', 'Chicago'],
    'Longitude': [-74.0060, -118.2437, -87.6298],
    'Latitude': [40.7128, 34.0522, 41.8781]
})

# Creating geometry column with lambda function
df['geometry'] = df.apply(lambda row: Point(row['Longitude'], row['Latitude']), axis=1)

# Converting DataFrame to GeoDataFrame
gdf = gpd.GeoDataFrame(df, geometry='geometry')
print(gdf)

Output:

          City  Longitude  Latitude                    geometry
0     New York    -74.0060   40.7128  POINT (-74.006 40.7128)
1  Los Angeles   -118.2437   34.0522 POINT (-118.2437 34.0522)
2      Chicago    -87.6298   41.8781  POINT (-87.6298 41.8781)

The lambda function provides a concise and elegant way to apply complex operations row-wise on the DataFrame. The lambda function in this case creates a new column ‘geometry’ which consists of Point objects generated from ‘Longitude’ and ‘Latitude’, facilitating the easy creation of the GeoDataFrame.

Bonus One-Liner Method 5: GeoPandas’ points_from_xy()

GeoPandas provides a utility function points_from_xy() that creates a geometry series of Points from two numeric columns. This one-liner is highly efficient for converting pandas DataFrames that contain separate longitude and latitude columns into a GeoDataFrame with point geometries.

Here’s an example:

import pandas as pd
import geopandas as gpd

# Sample DataFrame with longitude and latitude
df = pd.DataFrame({
    'City': ['New York', 'Los Angeles', 'Chicago'],
    'Longitude': [-74.0060, -118.2437, -87.6298],
    'Latitude': [40.7128, 34.0522, 41.8781]
})

# Converting DataFrame to GeoDataFrame in one line
gdf = gpd.GeoDataFrame(df, geometry=gpd.points_from_xy(df['Longitude'], df['Latitude']))
print(gdf)

Output:

          City  Longitude  Latitude                    geometry
0     New York    -74.0060   40.7128  POINT (-74.006 40.7128)
1  Los Angeles   -118.2437   34.0522 POINT (-118.2437 34.0522)
2      Chicago    -87.6298   41.8781  POINT (-87.6298 41.8781)

This succinct code snippet illustrates the power of the points_from_xy() function. By directly passing the longitude and latitude columns from the DataFrame, it returns a GeoSeries of Point geometries used to create the GeoDataFrame with minimal code.

Summary/Discussion

  • Method 1: Direct Constructor. Straightforward and clear. Limited customization for initial geometry creation.
  • Method 2: set_geometry() Function. Flexible. Requires an existing geometry-like column or series.
  • Method 3: Custom Function with apply(). Highly customizable. Potentially verbose and slower for large datasets.
  • Method 4: Lambda Function. Clear and compact. Syntax may be less readable to those unfamiliar with lambda functions.
  • Method 5: points_from_xy() Utility. Fastest and most succinct. Limited to the creation of Point geometries.