Add framealpha argument for legend #1626
Conversation
|
Travis failure looks like a dud. This seems like a simple enough wrapper around Anybody else want to give a +1 or other feedback? Edit: Barring minor style issues below. |
| def test_framealpha(): | ||
| x = np.linspace(1,100,100) | ||
| y = x | ||
| plt.plot(x,y, label='mylabel', lw=10) |
There was a problem hiding this comment.
Could you add a space after the first comma? Like so: plt.plot(x, y, label='mylabel', lw=10).
|
Seems fine -- perhaps it's a needless proliferation of keyword arguments -- OTOH, I can see why this is a common thing someone would want to do. @pelson: do you want to ring in on this? |
|
Both ways to get transparency doesn't work well with shadow though since shadow option assume full opacity of the legend frame. |
|
Hmmm... should we fix shadow while we're at it, then? |
|
This is what it looks like. It looks funny but I'm not sure what I should expect.... plot(range(10),range(10),label='hey')
legend(shadow=True).get_frame().set_alpha(0.5) |
No complaints from me. |
|
I think the shadow issue is probably fine for now... I agree -- it's not clear what it should do instead. @piti118: can you please rebase this on master so we can hit that big green merge button? |
|
That does look funny to me, but I can understand why some might not think |
|
Rebased.. I think I did it right. |
|
Yep. Looks good. Merging. |
Add framealpha argument for legend
|
Hmmm... I think I didn't write the change log... |
|
Oh, thanks. I should have noticed it before merging. I should be able to merge that manually. No worries. |
|
just pull #1677 |
Straight from matplotlib recipe. I think frame alpha belongs to keyword argument like
rather than