are you aware your features datasets omitted the ID column so its being difficult to map train and test columns?

Yeah @ibrahim you omitted the ID column. Can you assist with Dataset link

Hello zendians, sorry for the late reply, I was a little bit busy.

The raw features dataset HAS NO ID COLUMN, the index is only the (latitude, longitude, time).

You can map it to the train or test column through cKDTree or any distance based method.

In the benchmark notebook, they used nearest neighbour to map the ground truth data with features data.

I hope it's clear now, @Lone_Wolf and @ChinonsoVicky