Rename Bunch keys #44
Conversation
skada/datasets/_base.py
Outdated
| @@ -232,8 +232,8 @@ def pack( | |||
| target = np.concatenate(ys) | |||
| sample_domain = np.concatenate(sample_domains) | |||
| return (data, target, sample_domain) if return_X_y else Bunch( | |||
There was a problem hiding this comment.
you should also change the variable nalmes in the for loop
There was a problem hiding this comment.
To avoid confusion with the X and y in the for loop, I reuse the names Xs and ys instead of data and target respectively
|
It seems those we mentioned in the docs as well (in the README file). |
|
Plus sample generator functions have documentation (not sure if they implement packing locally but worth checking out). |
|
The documentation for |
|
Same for Office31 datasets. |
|
It seems the |
| @@ -470,4 +470,7 @@ def _load_office31( | |||
| data.append(content.astype(loader_spec.dtype)) | |||
| files['data'] = np.vstack(data) | |||
There was a problem hiding this comment.
just set X and y here no need to add stuff and pop it
There was a problem hiding this comment.
We can't. The files object comes from the load_files function, which is a sklearn function. Thus the keys 'data' and 'target' exist from the beginning.
Thus at some point we would have to remove these keys to not have duplicates with the keys 'X' and 'y'.
Thus the pop methods
#25
DomainAwareDatasetdef