Spatial_pooler compute active column output equivalence (learn vs no learning)

Engineering NuPIC
1

Hey guys just one quick question.

Are the following always equivalent, by that I mean the active columns would be the same?

sp = SpatialPooler(…)

At the ith step…

sp.compute(input, False, active_columns1)
sp.compute(input, True, active_columns2)
active_columns1 == active_columns2 ?

I’m looking for ways for doing a “dry run” for the compute function.

Cheers

2

Yes, you will get the same result. On the next compute though, you might not.

3

I see thanks for the confirmation. I wanted the sp to stop at a particular state and to do this I needed a way to copy the sp at the desired ith step OR just do a dry run and stop.