CODEWITHSUNDEEP
pythonpandasdataframecoordinatesdistance
Luis Ferreira
Luis Ferreira

Reputation: 53

Pandas Latitude-Longitude: Calculate the distance between two points and select the nearest one

I have a dataframe with 15k rows with coordinates and the names of the locations (mainly businesses)

spot lat1 l1 place lat2 lon2
1 41,1808128356934 -8,53291034698486 A 41.146749 -8.613889
1 41,1808128356934 -8,53291034698486 B 41.146105 -8.609868
2 41,1491432189941 -8,61034202575684 A 41.146749 -8.613889
2 41,1491432189941 -8,61034202575684 B 41.146105 -8.609868

I've been trying to perform a subtraction between both coordinates to find the distance (in meters) and then select only the the spot and place with less distance between them.

This is the code I was trying to use:

df['X_diff'] = df['lon1'] - df['lon2']
df['Y_diff'] = df['lat1'] - df['lat2']
df['dist'] = np.linalg.norm(df[['X_diff', 'Y_diff']], axis=1)

I also realized that the type of coordinates were different.

spot     object
lat1     object
lon1     object
place    object
lat2     float64
lon2     float64
dtype: object

How do I find the distance and select only the nearest one?

Upvotes: 1

Views: 598

Answers (1)

Quang Hoang
Quang Hoang

Reputation: 150785

Your lat1 and lon1 are strings. Maybe you want to replace , with . and convert to float before you calculate the norm:

df[['lat1', 'lon1']] = df[['lat1','lon1']].apply(lambda x: pd.to_numeric(x.str.replace(',','.')) )

Upvotes: 1

Related Questions