quant :
I have the following df
id xx yy time
0 1 553343.041098 4.178420e+06 1
1 1 553343.069815 4.178415e+06 2
2 1 553343.069815 4.178415e+06 3
3 2 553343.950755 4.178415e+06 1
4 2 553341.343829 4.178410e+06 6
The xx and yy is the position of each id at a certain point in time.
I would like to create an extra column in this df, which will be the difference in distance from one point of time to another (going from the smallest value of time to the next bigger, to the next bigger one etc), within the id group.
Is there a Pythonic way of doing so ?
moys :
You can do like below.
I didn't do df['distance_meters'] because it is straight forward.
df['xx_diff']=df.groupby('id')['xx'].diff()**2
df['yy_diff']=df.groupby('id')['yy'].diff()**2
If you don't need ['xx_diff'] & ['yy_diff'] columns in your dataframe, you can directly use the code below.
df['distance']= np.sqrt(df.groupby('id')['xx'].diff()**2+df.groupby('id')['yy'].diff()**2)
Output
id xx yy time xx_diff3 yy_diff3 distance
0 1 553343.041098 4178420.0 1 NaN NaN NaN
1 1 553343.069815 4178415.0 2 0.000825 25.0 5.000082
2 1 553343.069815 4178415.0 3 0.000000 0.0 0.000000
3 2 553343.950755 4178415.0 1 NaN NaN NaN
4 2 553341.343829 4178410.0 6 6.796063 25.0 5.638800