Toolbox: DataFrame points to GeoSeries

Tutorial presents how to transform points as floats written in DataFrame into GeoSeries object.

Updates

• 2021-06-28: GeoPandas method geopandas.points_from_xy included in the tutorial due to the nice response from GeoPandas developers on Twitter (here).

Requirements

1. pandas to read csv files with float coordinates,
2. geopandas to create GeoSeries.

import geopandas as gpd
import pandas as pd

Proof of Concept

GeoPandas allows us to transform any list of points into Point() with geopandas.points_from_xy() method:

# Transformation of floats to the Point

lat = 51.5074
lon = 0.1278

point = gpd.points_from_xy([lat], [lon])

Function

# Transform sample

def lat_lon_to_point(dataframe, lon_col='longitude', lat_col='latitude', crs='WGS84'):
    """Function transform longitude and latitude coordinates into GeoSeries.
    
    INPUT:
    
    :param dataframe: DataFrame to be transformed,
    :param lon_col: (str) longitude column name, default is 'longitude',
    :param lat_col: (str) latitude column name, default is 'latitude',
    :param crs: (str) crs of the output geoseries.
    
    OUTPUT:
    
    :return: (GeoPandas GeoSeries object)
    """
    geometry = gpd.points_from_xy(dataframe[lon_col], dataframe[lat_col])
    geoseries = gpd.GeoSeries(geometry)
    geoseries.name = 'geometry'
    geoseries.crs = crs
    
    return geoseries

The important thing is that longitude is x coordinate and latitude is y coordinate. We must set coordinate reference system (CRS) to create valid GeoSeries. In most cases CRS of point measurements is WGS84 popularized by the Global Positioning Systems.

You may now work with your GeoSeries and use spatial methods and attributes of this object!

Requirements – old algorithm

1. pandas to read csv files with float coordinates,
2. geopandas to create GeoSeries,
3. shapely to transform floats into Point.

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

Proof of Concept – old algorithm

Two floats may be easily transformed into Point with shapely. List or tuple of two floats is passed into the Point class constructor:

# Transformation of floats to the Point

lat = 51.5074
lon = 0.1278

sample_coordinate = [lon, lat]  # x, y
point = Point(sample_coordinate)

Function – old algorithm

The hardest part of transformation is to transform DataFrame float columns at once. We use for it .apply() method and lambda function.

# Transform float columns into geoseries

def lat_lon_to_point(dataframe, lon_col='longitude', lat_col='latitude', crs='WGS84'):
    """Function transform longitude and latitude coordinates into GeoSeries.
    
    INPUT:
    
    :param dataframe: DataFrame to be transformed,
    :param lon_col: (str) longitude column name, default is 'longitude',
    :param lat_col: (str) latitude column name, default is 'latitude',
    :param crs: (str) crs of the output geoseries.
    
    OUTPUT:
    
    :return: (GeoPandas GeoSeries object)
    """
    geometry = dataframe.apply(lambda x: Point([x[lon_col], x[lat_col]]), axis=1)
    geoseries = gpd.GeoSeries(geometry)
    geoseries.name = 'geometry'
    geoseries.crs = crs
    
    return geoseries