Added documentation for the Selector object, as well as references to it #1841
Conversation
rcap107
left a comment
rcap107
left a comment
There was a problem hiding this comment.
Thanks for the PR @emassoulie! I added some comments on the format of the docs.
I think the user guide should also be updated to explain the use of expand and expand_index.
|
|
||
| def all(): | ||
| """Select all columns. | ||
| """Select all columns. See also :class:`skrub.selectors.Selector()`. |
There was a problem hiding this comment.
I'm not sure anymore there should be a "See also" referring to Selector for all these functions, especially since we are saying that the Selector should not be used by the users.
However, we might want to use the See Also field to refer to the user guide, where we explain that the function is returning a selector and that the selector has expand and expand_index.
This also means that we need to add a proper mention of this in the user guide, with examples and explanation.
Also note the formatting of the See Also: it's a specific field in the numpy docs. It's used in various places in the library, like the StringEncoder.
| Parameters | ||
| ---------- | ||
| df : dataframe | ||
There was a problem hiding this comment.
There should be an example here showing what expand does when applied to a dataframe.
| def expand(self, df): | ||
| """Lists the column names that the selector would retain if applied to | ||
| the dataframe `df`. | ||
There was a problem hiding this comment.
Either here or in the examples there should be a message mentioning that this can be useful for using the selectors with standard dataframe libraries.
| The list of `df`'s columns that the item would select. In effect, | ||
| running `df[sel.expand(df)]` should give the exact same result as | ||
| `sel.transform(df)`. |
There was a problem hiding this comment.
The lines I removed here are a better fit for the Notes section, or for an example.
| The list of `df`'s columns that the item would select. In effect, | |
| running `df[sel.expand(df)]` should give the exact same result as | |
| `sel.transform(df)`. | |
| The list of `df`'s columns that the item would select. |
| def expand_index(self, df): | ||
| """Lists the indices of dataframe `df`'s columns that the selector | ||
| would retain if applied to `df`. | ||
There was a problem hiding this comment.
The comments I made for the expand case also apply here
2 participants
Closes #1791 .
In order to give proper documentation to the Selector class's
expandandexpand_indexmethod, added a doc entry to the class itself (which was already public) and references to that documentation in the docstring of each of its inherited classes.