SpatialPooler::inhibitColumnsGlobal_ comments, unit tests, cleanup

[ctrl-z-9000-times]

I hope these comments help make it clear how the new SpatialPooler::inhibitColumnsGlobal_ method works. Also added 3 unit test cases for this method.

Additional fixes to #80

[breznak]

💯 perfect! This is much more understandable and useful for the future.
Thank you for doing the final touches!

[breznak]

@ctrl-z-9000-times

[ RUN ] SpatialPoolerTest.testInhibitColumnsGlobal
Too long with no output (exceeded 10m0s)

Something stalls the tests for 10+mins and the CI then kills the build

[breznak]

Reopening to restart the CI, maybe it helps

[ctrl-z-9000-times]

Maybe you you could take just the commit with the documentation. I wrote those unit tests without realizing that there are already some decent unit tests for that method. That method is also tested by testing the compute method with global inhibition, since that will call this method.

[breznak]

I was trying to figure why those tests get stuck, but couldn't find it..this is also a solution

[dkeeney]

That ConnectionsPerformance test never seems to complete. Well, I guess it did complete... ConnectionsPerformanceTest (127469 ms total) Also, the HelloSPTPTest test. HelloSPTPTest.performance (26422 ms) Considering that the test machine is probably a rather large processor, that is a long time. @breznak, can you look at that and see if there is a way to take less resources?

…

On Fri, Nov 9, 2018 at 3:23 PM breznak ***@***.***> wrote: ***@***.**** approved this pull request. I was trying to figure why those tests get stuck, but couldn't find it..this is also a solution — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#84 (review)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AFBa_0dmjUzAkodRJ_HGw5PGmiZ0QfR0ks5utg53gaJpZM4YUxZF> .