Open zohaad opened 4 years ago
Hey, thank you for the suggestion. Actually, I need to do more than just removing shuffling to look into this since each company is loaded one after the other into the data set. Will let you know.
You also need to add a date < filingDate
restriction to the split
In
trn_NetCompustat.py
you need to pass inshuffle=False
tosklearn.model_selection.train_test_split