CODEWITHSUNDEEP
pythonsetup.py
Rose Perrone
Rose Perrone

Reputation: 63546

TypeError: setup() got multiple values for keyword argument 'packages'

This error occurs on line 31 as noted below when I run the following command according to these instructions. How am I getting multiple values for packages?

python setup.py config_fc --fcompiler=gnu95 \
         --f77flags='-fdefault-real-8' \
         --f90flags='-fdefault-real-8' build

.

from numpy.distutils.misc_util import Configuration
from numpy.distutils.system_info import get_info
import os, sys

import sys

fflags= '-fdefault-real-8 -ffixed-form'

# TODO: Fix it so that these flags are default.

config = Configuration(
    'glmnet',
    parent_package=None,
    top_path=None
)


f_sources = ['glmnet/glmnet.pyf','glmnet/glmnet.f']

config.add_extension(name='_glmnet',sources=f_sources)
config_dict = config.todict()
if __name__ == '__main__':
    from numpy.distutils.core import setup
    setup(version='1.1-5',
          description='Python wrappers for the GLMNET package',
          author='David Warde-Farley',
          author_email='[email protected]',
          url='github.com/dwf/glmnet-python',
          license='GPL2',
          requires=['NumPy (>= 1.3)'],
          packages=['glmnet'],            ### LINE 31
          **config_dict)

Upvotes: 1

Views: 695

Answers (1)

Martijn Pieters
Martijn Pieters

Reputation: 1121744

The Configuration(package_name).to_dict() result already includes the packages=[package_name] entry.

By including it in setup() and in the config_dict mapping, setup() is supplied that keyword twice. You could remove it from the setup() call (line 31):

setup(version='1.1-5',
      description='Python wrappers for the GLMNET package',
      author='David Warde-Farley',
      author_email='[email protected]',
      url='github.com/dwf/glmnet-python',
      license='GPL2',
      requires=['NumPy (>= 1.3)'],
      **config_dict)

I see that the author of the package added the packages line explicitly 4 years ago; I strongly suspect that the NumPy behaviour has since changed.

You should probably file a another issue with the project.

Upvotes: 2

Related Questions