Python script fails with coordinate variables (lat, lon) due to argument 2 of type 'double'

Question

I am plotting points using ogr in Python (Python 3!), and it works fine if I input my coordinates into the code as numbers, but the following code fails if I try and use the coordinates from a csv file.

Working code:

from osgeo import ogr

point = ogr.Geometry(ogr.wkbPoint)
point.AddPoint(40.729047, -73.957472)
print(point.ExportToWkt())

Code that breaks:

from osgeo import ogr

datatest = pd.read_csv('data.csv')

p1_lat, p1_lon = datatest['POINT1_LAT'], datatest['POINT1_LON']

point = ogr.Geometry(ogr.wkbPoint)
point.AddPoint(p1_lat, p1_lon)
print(point.ExportToWkt())

That second example fails with the error TypeError: in method 'Geometry_AddPoint', argument 2 of type 'double'

What am I doing wrong and how can I fix it so that the code can call coordinates from my csv file? Thanks.

EDIT: as requested, if I print the two variables I get:

0    40.729047
Name: POINT1_LAT, dtype: float64
0   -73.957472
Name: POINT1_LON, dtype: float64

EDIT2: This is Python 3

Answer 1

You are adding pandas dataframes with AddPoint instead of floats.

import pandas as pd

pts = pd.DataFrame({'p': [3.4532], 'q' : [5.674]})

print(pts['p'])
print(pts['q'])

print(float(pts['p'][0]))
print(float(pts['q'][0]))

0    3.4532
Name: p, dtype: float64
0    5.674
Name: q, dtype: float64
3.4532
5.674

This should work: -

from osgeo import ogr

datatest = pd.read_csv('data.csv')

p1_lat, p1_lon = datatest['POINT1_LAT'], datatest['POINT1_LON']
p1_lat = float(p1_lat[0])
p1_lon = float(p1_lon[0])

point = ogr.Geometry(ogr.wkbPoint)
point.AddPoint(p1_lat, p1_lon)
print(point.ExportToWkt())