Call _transform_vmin_vmax during SymLogNorm.__init__

[Salganos]

A fix for #6750

[tacaswell]

👍

Can you also add a test? It does not need to be an image test, just the minimal code that would have raised before.

[Salganos]

Not sure where to add that, but I'll look into it. The minimal test should be something like:
cb.ColorbarBase(plt.figure().add_subplot(111), norm = clr.SymLogNorm(linthresh = .1, vmin = -1, vmax = 1))

Edit: I think I've found the spot.

[tacaswell]

This should be it's own test (any function name that starts with test_ will work).

It also needs the @cleanup decorator.

[tacaswell]

❤️ pep8/pycodestyle

There is some trailing white space:

FAIL: matplotlib.tests.test_coding_standards.test_pep8_conformance_installed_files
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/home/travis/build/matplotlib/matplotlib/venv/lib/python3.5/site-packages/nose/case.py", line 198, in runTest
    self.test(*self.arg)
  File "/home/travis/build/matplotlib/matplotlib/lib/matplotlib/tests/test_coding_standards.py", line 244, in test_pep8_conformance_installed_files
    expected_bad_files=expected_bad_files)
  File "/home/travis/build/matplotlib/matplotlib/lib/matplotlib/tests/test_coding_standards.py", line 143, in assert_pep8_conformance
    assert_equal(result.total_errors, 0, msg)
AssertionError: 1 != 0 : Found code syntax errors (and warnings):
/home/travis/build/matplotlib/matplotlib/lib/matplotlib/tests/test_colors.py:226:1: W293 blank line contains whitespace

This may seem petty and annoying, but keeping a consistent style (including no trailing white space) really does help on big projects with many contributors.

[Salganos]

Whoops. The perils of not using a proper development environment.

[NelleV]

👍 LGTM

[efiring]

Apart from the minor suggestion to delete a line from the test, this looks good--it makes sense.

[efiring]

I think this close is unnecessary; the @cleanup decorator handles it, doesn't it?

[tacaswell]

It does, but this does not hurt

[tacaswell]

Changed my mind about targeting 2.x . If someone disagrees, I will not protest to it being backported.

[efiring]

It's not critical, but I am mildly in favor of the backport. The change has no effect in most uses, but blocks a potential failure mode.

[QuLogic]

@efiring Are you going to backport?

[efiring]

Sure, I can do that in a few minutes.

[efiring]

Backported to v2.x as 0ab4599