Error in fisher.test(table(predictions > 0.5, labels.train.cv.test))

[ashokpatowary]

Hi @andredsim

I am trying to run bambu in my dataset (~250M reads over 10 samples). Initially I ran with quant = FALSE and it finished without any error. After that I tried with quant = TRUE and continuously encountering errors. Even running with quant = FALSE throwing error. Can you please looks into this and suggest?

Thanks

pValue: 0.106488269794721
AUC: 0.894230769230769
prediction accuracy (CV) (higher for splice donor than splice acceptor)
Error: BiocParallel errors
  1 remote errors, element index: 6
  2 unevaluated and other errors
  first remote error:
Error in fisher.test(table(predictions > 0.5, labels.train.cv.test)): 'x' must have at least 2 rows and columns
In addition: There were 50 or more warnings (use warnings() to see the first 50)

[lingminhao]

Hi @ashokpatowary ,

May I ask if you use an annotation file that is much smaller than the normal annotation file (e.g. the human) during your second run ?

If you managed to do the first run (even with quant = FALSE) successfully, you should not encounter any issue with quant = TRUE later as quantification is not related to the use of fisher test. It may be possible that the data / the argument used is different between the runs.

Best,
Min Hao