Skip to content
This repository has been archived by the owner on Sep 1, 2023. It is now read-only.

NuPIC Needs Integration Test Illustrating Correct Output To Use #2684

Open
cogmission opened this issue Oct 20, 2015 · 2 comments
Open

NuPIC Needs Integration Test Illustrating Correct Output To Use #2684

cogmission opened this issue Oct 20, 2015 · 2 comments

Comments

@cogmission
Copy link
Contributor

@rhyolight @subutai @scottpurdy @chetan51 @breznak

There needs to be an example developers can refer to which illustrates how interaction between the algorithms is accomplished.

This issue author is proof of that need.

Being the most official porter of NuPIC code, and therefore the most objective test for correct interpretation of NuPIC, the fact that I had no reference and mistakenly used the incorrect output when utilizing and implementing the interaction between algorithms - is proof of this need.

  • The test should show getting data from Encoders, SpatialPoolers, TemporalMemories, Classifiers, Anomaly(s), AnomalyLikelihood(s) and feeding that data into the appropriate subsequent algorithm or algorithm(s) in the case where output can be sent into multiple subsequent algorithm types.
@breznak
Copy link
Member

breznak commented Oct 20, 2015

👍 I've burned my fingers when getting some TP output and passing it manually to Anomaly!

@cogmission
Copy link
Contributor Author

@breznak I agree, however I think that mentioning that here may be distracting. It's probably better to mention this Issue from #2536 rather than have a discussion of that here. This issue should probably remain pure so that it doesn't get coopted by off-topic conversation (as I have been guilty of doing too!)

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants