whaleloops / KEPT

auto icd coding with prompt
MIT License
46 stars 17 forks source link

MIMIC-III-rare50 labels #5

Open JosephSBoyle opened 1 year ago

JosephSBoyle commented 1 year ago

Hi there, how can I obtain a set of labels for the rare50 dataset?

In collectrare50data.py, taking the union of the train / dev / test labels gives 56 labels - is this correct?

dev_labels_set   = set(dev_labels)
train_labels_set = set(train_labels)
test_labels_set  = set(test_labels)
print(len(train_labels_set | dev_labels_set | test_labels_set))
# outputs '56'
exit()

Thanks for your help.