Wrap grd2cpt

[willschlitzer]

Wraps the GMT function grd2cpt. Code and documentation borrowed heavily from grdcut and makecpt.

[willschlitzer]

For reasons I'm not sure, I'm getting a GMTImageComparisonFailure: images not close error when I run test_grd2cpt_grdimage. I'm confused because the code for both fig_ref and fig_test are identical (as I copy and pasted to make sure I didn't have a typo). When I ran the same code in a Jupyter notebook, I also had images that looked very different. Any idea what is causing this inconsistency?

[seisman]

I haven't checked it carefully, but it may be related to #372

[seisman]

I'm OK with the wrapper and the aliases, but most of the tests don't make sense to me.

For example, in the following test, if cmap="rainbow" doesn't work as expected (although unlikely as the wrapper is very simple), it would affect both the reference and test images, thus the tests always pass.

@check_figures_equal()
def test_grd2cpt_set_cpt(grid):
    """
    Test function grd2cpt to create a CPT based off a grid input and a set CPT.
    """
    fig_ref, fig_test = Figure(), Figure()
    # Use single-character arguments for the reference image
    fig_ref.basemap(B="a", J="W0/15c", R="d")
    grd2cpt(grid=grid, cmap="rainbow")
    fig_ref.colorbar(B="a2000")
    fig_test.basemap(frame="a", projection="W0/15c", region="d")
    grd2cpt(grid=grid, cmap="rainbow")
    fig_test.colorbar(frame="a2000")
    return fig_ref, fig_test

[willschlitzer]

I'm fine with deleting tests. I was mostly copying the format of test_makecpt.py, since that seems like the most similar module.

[seisman]

I'm fine with deleting tests. I was mostly copying the format of test_makecpt.py, since that seems like the most similar module.

Some of the makecpt tests make sense because they're using @pytest.mark.mpl_image_compare. I think we need to think more about these tests.

@weiji14 As you wrote the makecpt wrapper, I'll let you decide what to do.

[weiji14]

Just a few minor comments, getting close!

[weiji14]

Great! Thanks for your patience @willschlitzer, and also thank you @seisman for the comprehensive review!