search

CAHLR / pyBKT

Python implementation of Bayesian Knowledge Tracing and extensions
MIT License
201 stars 63 forks source link

Split train test data at sequence level #47

Open capriceZ opened 1 year ago

capriceZ commented 1 year ago

Hello!

I have been experimenting with running model using data split at sequence level rather than student level. But whenever I set 'multigs'=true, I can fit the model but I can NOT evaluate the model using model.evaluate(). I got the following error: File "/lib/python3.9/site-packages/pyBKT/util/data_helper.py", line 275, in convert_data data_temp[data_ref[i]][i] = data[i] IndexError: index 1 is out of bounds for axis 0 with size 1

All other model configs work fine for fitting and evaluating. Only models with 'multigs'=true have this issue. Can you help me understand what happened?

Thank you!

abadrinath947 commented 12 months ago

Can you please provide more information about what commands you ran and what dataset you used? You do not need to share the dataset if you are unable, but if you are able to replicate it across a public dataset (such as AS/CT), that would be helpful!