bug in Pandas when reading files with 'Infinity'

[githeap]

By default this fork of QtPlaskin uses pandas.read_csv function to speed up data loading.
ZdPlaskin is a Fortran code and uses "Infinity" to represent inf values in output text files.
There is an issue #10065 in Pandas with reading such files. Pandas read_csv interprets "Infinity" as a string, not a floating point value.

Untill the issue #10065 is not resolved in Pandas, as a workaround, we fallback to old (slower) data import for datasets containing "Infinity"s. a5b1559

Update: Add a very straightforward fix to Pandas #28181. Hope, someone will come up with smarter solution.

• Get ENH: Enable read_csv interpret 'Infinity' as floating point value #10065 pandas-dev/pandas#28181 be merged
• Wait for Pandas release
• Fix tests (remove xfail)
• Update setup.py

  qtplaskin/setup.py

  Lines 48 to 56 in 1aeb4a7

  	install_requires=[
  	'future', # for builtins
  	'numpy',
  	'scipy',
  	'matplotlib',
  	'h5py',
  	'mpldatacursor',
  	# 'pyqt' # cannot be installed through pip. Install PyQt5 manually
  	],

[erwanp]

Well done on getting it approved!

[githeap]

pandas-dev/pandas#28181 was merged in Pandas 1.0