[ENH] improve performance for polars' complete (#1389)

* remove ambiguity with structs * fix docs * pin numpy * fix docs * fix docs * add polars.expand * add polars.expand * add tests polars expand * pin setuptools * fix docs * fix docs * add extra test for polars expand * remove assumptions when fill_value is a dict * add test for pl.Series * fix column selection logic --------- Co-authored-by: samuel.oranyeli <[email protected]> Co-authored-by: Eric Ma <[email protected]>
pyjanitor-devs · Jul 31, 2024 · 633c891 · 633c891
1 parent a14061c
commit 633c891
Show file tree

Hide file tree

Showing 6 changed files with 421 additions and 88 deletions.
diff --git a/environment-dev.yml b/environment-dev.yml
@@ -25,7 +25,7 @@ dependencies:
   - mypy
   - natsort
   - numba
-  - numpy
+  - numpy==1.24.4
   - openpyxl
   - pandas-flavor
   - pandas-vet
@@ -51,3 +51,5 @@ dependencies:
   - xorg-libxrender
   - pip:
       - mknotebooks
+  # Temporarily pinned to fix CI
+  - setuptools==70.3.0
diff --git a/janitor/functions/convert_date.py b/janitor/functions/convert_date.py
@@ -102,7 +102,6 @@ def convert_matlab_date(
     return df.assign(**dictionary)
 
 
-@pf.register_dataframe_method
 @pf.register_dataframe_method
 @refactored_function(
     message=(

diff --git a/janitor/polars/__init__.py b/janitor/polars/__init__.py
@@ -1,5 +1,5 @@
 from .clean_names import clean_names, make_clean_names
-from .complete import complete
+from .complete import complete, expand
 from .dates_to_polars import convert_excel_date, convert_matlab_date
 from .pivot_longer import pivot_longer, pivot_longer_spec
 from .row_to_names import row_to_names
@@ -10,6 +10,7 @@
     "clean_names",
     "make_clean_names",
     "row_to_names",
+    "expand",
     "complete",
     "convert_excel_date",
     "convert_matlab_date",