feat: implement cols_reorder() for flexible column reordering
#625
Conversation
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## main #625 +/- ##
==========================================
+ Coverage 90.85% 90.86% +0.01%
==========================================
Files 46 46
Lines 5423 5432 +9
==========================================
+ Hits 4927 4936 +9
Misses 496 496 ☔ View full report in Codecov by Sentry. |
great_tables/_spanners.py
Outdated
| function: `cols_reorder()`. By passing a list of column names to the `columns=` parameter, | ||
| the table will be reordered accordingly: | ||
| ```{python} | ||
| col1, col2, col3, *cols = exibble.columns # col1="num", col2="char", col3="fctr" |
There was a problem hiding this comment.
This is a really good example!
There was a problem hiding this comment.
I've renamed the variables to use their actual column names instead of generic names like col1. This should improve readability.
|
|
||
| if isinstance(columns, str): | ||
| columns = [columns] | ||
|
|
There was a problem hiding this comment.
I wonder if we should raise here, if only to provide a more user-friendly message. The user will instead see "Reordering vars must contain all boxhead vars." if not providing all columns, which is good for a developer-facing error message but it exposes some of the underlying implementation that the user might not know about.
There was a problem hiding this comment.
Great suggestion! Should we raise the exception inside the function or modify the error message to be more user-friendly rather than developer-focused?
There was a problem hiding this comment.
Good question. I’m for slightly tweaking the wording in the existing location. Just switch vars with columns and avoid the use of boxhead.
There was a problem hiding this comment.
How about "Column reordering must include all columns in the table."?
Hello team,
This PR introduces
cols_reorder(), allowing users to reorder columns in a specified order easily.While the existing
move_*()functions are useful for repositioning a few columns, they don't provide fine-grained control over column order. I've often had to rely onPandasorPolarsfor this task. The implementation of this function follows a straightforward approach, similar to the existingmove_*()functions.Feedback is always welcome!😊