Converting x/y in Pandas DataFrame to lat/long

Question

I have written the following Python code to convert various input files into a shapefile using Pandas.

The coordinates of the input file are x,y which is what my shapefile outputs to, but I would like to add in some code which converts the input file to lat/long.

However, I'm unsure how to go about this

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


input_file = "...input_file.csv" 
file_extension = os.path.splitext(input_file)[-1].lower()
output_file = input_file[:-4]


if file_extension == ".xyz" or ".asc":
    df  = pd.read_table(input_file, skiprows=2, sep=r'\,|\t', engine='python', names=['x', 'y', 'z'])
    df.columns = ["x", "y", "z"]

elif file_extension == ".txt" or ".text" or ".csv":
    df = pd.read_csv(input_file, sep='\,|\t')
    df.columns = ["x", "y", "z"]
        

gdf = gpd.GeoDataFrame(df, geometry=df.apply(lambda row: Point(row.x,row.y,row.z), axis=1,), crs='EPSG:4326')

gdf.to_file(f"{output_file}.shp") 
shapefile = f"{output_file}.shp"

print("Shapefile Created!")

Answer 1

Firstly I recommend the os.path.splitext() function from the os.path module to get file_extension and output_file in a bit more elegant way:

from os.path import splitext

input_file = '.../input_file.csv' 
output_file, file_extension = splitext(input_file)
print(output_file, file_extension)

Let's assume there is a .csv-file called 'points.csv', see image below:

On this stage the GeoDataFrame will be created:

import pandas as pd
import geopandas as gpd
from os.path import splitext

absolute_path_to_file = '.../input_file.csv'

file_name, file_extension = splitext(absolute_path_to_file)

if file_extension == ".txt" or ".text" or ".csv":
    df = pd.read_table(absolute_path_to_file, sep=',')

gdf = gpd.GeoDataFrame(df, geometry=gpd.points_from_xy(df['x'], df['y'], df['z']), crs='EPSG:25832')

print(gdf)

that results in (keep in mind that the point's geometry representation is POINT Z):

   id          x           y      z                                  geometry
0   1  513072.05  5402445.83  250.0  POINT Z (513072.050 5402445.830 250.000)
1   2  513212.88  5402852.80  245.0  POINT Z (513212.880 5402852.800 245.000)
2   3  513733.74  5403658.99  239.0  POINT Z (513733.740 5403658.990 239.000)

Now it is possible to get latitude, longitude, and altitude using:

• the GeoPandas's to_crs() function as was suggested by @user2856:

   import pandas as pd
   import geopandas as gpd
   from os.path import splitext
   from pyproj import CRS
  
   absolute_path_to_file = '.../input_file.csv'
  
   file_name, file_extension = splitext(absolute_path_to_file)
  
   if file_extension == ".txt" or ".text" or ".csv":
       df = pd.read_table(absolute_path_to_file, sep=',')
  
   gdf = gpd.GeoDataFrame(df, geometry=gpd.points_from_xy(df['x'], df['y'], df['z']), crs='EPSG:25832')
  
   crs = CRS.from_string('EPSG:4326')
   gdf = gdf.to_crs(crs=crs)
  
   print(gdf)
  
      id          x           y      z                              geometry
   0   1  513072.05  5402445.83  250.0  POINT Z (9.17792 48.77488 250.00000)
   1   2  513212.88  5402852.80  245.0  POINT Z (9.17985 48.77854 245.00000)
   2   3  513733.74  5403658.99  239.0  POINT Z (9.18697 48.78578 239.00000)
  

• Pandas and a custom function with the pyproj's Transformer class:

    import pandas as pd
    from os.path import splitext
    from pyproj import CRS, Transformer
  
    def xyz_to_latlonalt(x, y, z, crs_in, crs_out):
        crs_from = CRS.from_user_input(crs_in)
        crs_to = CRS.from_user_input(crs_out)
        proj = Transformer.from_crs(crs_from, crs_to, always_xy=True)
        coordinates = proj.transform(x, y, z)
        return coordinates
  
    absolute_path_to_file = '.../input_file.csv'
  
    file_name, file_extension = splitext(absolute_path_to_file)
  
    if file_extension == ".txt" or ".text" or ".csv":
        df = pd.read_table(absolute_path_to_file, sep=',')
  
    df['lat'] = df.apply(lambda row: xyz_to_latlonalt(row['x'], row['y'], row['z'], 25832, 4326)[0], axis=1)
    df['lon'] = df.apply(lambda row: xyz_to_latlonalt(row['x'], row['y'], row['z'], 25832, 4326)[1], axis=1)
    df['alt'] = df.apply(lambda row: xyz_to_latlonalt(row['x'], row['y'], row['z'], 25832, 4326)[2], axis=1)
  
    df = df.drop(['x', 'y', 'z'], axis=1)
  
    print(df)
  
       id       lat        lon    alt
    0   1  9.177920  48.774878  250.0
    1   2  9.179850  48.778536  245.0
    2   3  9.186966  48.785778  239.0