'Spatial Point Pattern Analysis Python

I am observing grid patterns in my raw location data and am trying to clean such locations without getting rid of a lot of valid locations or any other pattern that is common which I can filter out.

As I am new to location data, I am not sure what are the methods I could explore to get rid of these data points. I will first try to look into DBSCAN, but am hoping to seek more advice on what other methods I can explore. The dataset is quite large (petabytes) so I have to keep in mind the scalability issue. I tried searching up on spatial point analysis patterns but I am not able to find anything that might help.