Features wrongly generated

Hi, we’re currently having an issue where we wrongly generate features which looks something like:

This is caused because the labels for the datapoints are written incorrectly causing data to be labeled randomly. We’ll be fixing this shortly!

Issue has been fixed. However, because we cache the DSP result this issue might persist in your account until the cache is invalidated. If you see this you can invalidate the cache by:

  1. Changing the label of a sample (just change it and change it back again).
  2. Or by removing the signal processing block and re-adding it.

If you have any questions, let me know on this thread and we’ll assist you.