[Bug 242672] science/py-scikit-learn: Fails due to unresolved symbols (py36-scikit-learn-0.20.3_1)

[bugzilla-noreply at freebsd.org]

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=242672

--- Comment #11 from Andy Mender <andymenderunix at gmail.com> ---
(In reply to Andy Mender from comment #10)

My apologies! I just now realized `cblas` and `blas` are NOT the same thing.

I tested science/py-scikit-learn-0.20.3 again, this time linked against blas
(not cblas!) via `LDFLAGS=-lblas`and it DID in fact work. I got the following
warnings while running my SVM model, but it still generated the expected
results:
```
/usr/local/lib/python3.6/site-packages/sklearn/utils/validation.py:595:
DataConversionWarning: Data with input dtype int64 was converted to float64 by
StandardScaler.
  warnings.warn(msg, DataConversionWarning)
/usr/local/lib/python3.6/site-packages/sklearn/utils/validation.py:595:
DataConversionWarning: Data with input dtype int64 was converted to float64 by
StandardScaler.
  warnings.warn(msg, DataConversionWarning)
/usr/local/lib/python3.6/site-packages/sklearn/utils/validation.py:595:
DataConversionWarning: Data with input dtype int64 was converted to float64 by
StandardScaler.
  warnings.warn(msg, DataConversionWarning)
```

I also suggest this change to the LIB_DEPENDS in the Makefile to make it depend
on math/blas, instead of math/cblas like it did before:
`LIB_DEPENDS=   libblas.so.2:math/blas`

-- 
You are receiving this mail because:
You are the assignee for the bug.