README.rst | 2 1 + 1 - 0 !
1 file changed, 1 insertion(+), 1 deletion(-)

 fix tests for current scipy
 scipy has made changes to its class hierarchy, update the tests to match

README.rst | 4 2 + 2 - 0 !
tests/test_dataframe_mapper.py | 8 4 + 4 - 0 !
2 files changed, 6 insertions(+), 6 deletions(-)

---

sklearn_pandas/pipeline.py | 3 1 + 2 - 0 !
1 file changed, 1 insertion(+), 2 deletions(-)

 [patch] drop use of six

This package has an undeclared requirement on six, but since Python 2 is
long unsupported, remove it and just use the Python 3 idiom.

sklearn_pandas/pipeline.py | 7 5 + 2 - 0 !
tests/test_pipeline.py | 70 70 + 0 - 0 !
2 files changed, 75 insertions(+), 2 deletions(-)

 fix importerror with sklearn 1.7.0+ by replacing tosequence

- Remove deprecated sklearn.utils.tosequence import (removed in sklearn 1.7)
- Convert steps to list at start of __init__ to handle all input types
- Replace tosequence(steps) with direct list assignment
- Fixes compatibility with scikit-learn >= 1.7.0
- Maintains backward compatibility with older sklearn versions
- Add regression tests for list, tuple, and generator inputs
- Verify steps attribute is always a list type

The tosequence utility was deprecated in sklearn 1.5 and removed in 1.7.
Using list() directly provides the same functionality without depending
on sklearn internals.

Fixes #267