Bug in ridge regression with induct_used=1?

ciuccislab / pyDRTtools

An intuitive python GUI to compute the DRT

MIT License

61 stars 27 forks source link

Bug in ridge regression with induct_used=1? #7

Closed ESaundersNano closed 8 months ago

ESaundersNano commented 11 months ago

Thank you for the great module. I have noticed that if induct_used=1 when calling simple_run in a notebook or by selecting "with inductive element" in the GUI, there is an issue forming the G matrix and an error is caused.

ESaundersNano commented 11 months ago

This issue is at least apparent if the example dataset from the tutorial is used, but also happens when I use my data.

ciuccislab commented 9 months ago

The issue arises from matrix mismatch in a line of code that enforces non-negativity constraint. The issue has been fixed in the line of code below. Nonetheless, a new version of tutorial will be released with the new/latest version of pyDRTtools. You can run main GUI, there will be no issue again. Thank you.