Repository navigation
Getting non-encoded dataset #408
Description
If I import, for instance, the adult dataset: https://www.openml.org/d/1590
get_data automatically encodes categorical features:
dataset = oml.datasets.get_dataset(1590) # Adult dataset
X, y, attribute_names = dataset.get_data(target=dataset.default_target_attribute,
return_attribute_names=True)
data = pd.DataFrame(X, columns=attribute_names)
display(data.head())
->'education' is encoded as numbers
Is there a way to get the original non-encoded features, so I can look at them with pandas?
Thanks!
Activity
mfeurer commentedon Feb 26, 2018
No, there isn't. I wrote that part before knowing how nicely pandas could do the job here. That would be a great feature! Related to #364 and #257.
mfeurer commentedon Mar 7, 2019
Should be possible with the latest merge of #548. Please reopen if this is not the case.