Open sneumann opened 6 years ago
Hi Steffen,
could you provide me a minimal example? One thing I notice is that you seem to be applying the batch correction row-wise, but that may have to do with the data (if your mass peaks are in the rows then your approach is correct). Second thing is that you seem to try to correct raw counts. I would take logs, definitely. Finally: it may happen that tobit regression fails to converge. Who knows you hit one of these cases. Again, happy to have a closer look.
Cheers, Ron
Hi Ron & Jos,
we're running into a numeric error with:
(see below for the full traceback).
least-square
androbust regression
compute without error, butcensored regression
fails as above. We unsure how to debug which parts of our input might be the cause, and whether it is really in the data or we passed the wrong inputs. Any hints appreciated.Yours, Steffen and Pauline