How to handling NaN values while creating a geojson

Im trying to create a .geojson file from DataFrame. I haver this table:

        long       lat  elev
0  40.392987 -3.652552   0.0
1  41.471062  2.041841   0.0
2  41.410016  2.175180   0.0
3        NaN       NaN   NaN

And this is the function for create .geojson file

    def data2geojson(df):
        features = []
        df.apply(lambda X: features.append(
            geojson.Feature(geometry=geojson.Point((X["lat"],
                                                    X["long"],
                                                    X["elev"])),
                            properties={"country": "Spain"})), axis=1)
        with open('map.geojson', 'w') as fp:
            geojson.dump(geojson.FeatureCollection(features), fp, sort_keys=True)

This code works if I only select the first three columns but throws the next error: Out of range float values are not JSON compliant: nan

I dont know how to handle this more than with try: and except ValueError:

But throwing an exception break the function and cant continue if there are more values...How can I handle this??

Solution

This error comes from the fact that JSON does not accept NaN values (which are considered float in Python by the way). Before converting it to geojson you then need to edit these values in your input dataframe.

Depending on what you want to do with NaN values, you can fix this error with the following methods:

You can drop them all with the following instruction. This will remove the row as soon as at least one of the value of the row is NaN:

df = df.dropna()

You can replace them by any value. The following replaces any NaN by 0:

df = df.fillna(value=0)