subplot legend consistent legend location matplotlib

SUBPLOT LEGEND CONSISTENT LEGEND LOCATION MATPLOTLIB FULL

I agree that we should come up with a better solution. Let's have some more examples for/against this change. Although this complex multiplot doesn't look good, simpler 1-4 subplot figures tend to look better I feel. Though, you can't be certain if people on average don't like the current behavior.Īll in all, one example is not enough of a justification for the revert. Until anyone finds a better solution - which, given the development timeline of Plots, could take ages - and since it is clearly affecting users, I would propose, again, to revert the PR. As you said, this thread is all about opinions.

If you have the will and energy, then go ahead collect some data, and present facts. It's hard to tell which one people like more without a survey and proper methodology. Well, I did this with approval, we had a Plots maintainers call with and just sad that I'm being accused here.Īnd this is your opinion too. We can then work on a draft PR to solve the scaling issue. It's good that the change triggered the discussion, now we should wait for someone to come up with a better solution. Again, the question is: which is more important in a plot, data or decorations ? My argument is that decorations should not hide data (and others seem to share this opinion). On the GR plot, the data is unreadable, hidden by the legend box (which was the initial argument against these changes). On this particular PyPlot example, the subplots are bigger, and the data can be interpreted. Making plots look like what you get in matplotlib is not and should not be the development guideline.

SUBPLOT LEGEND CONSISTENT LEGEND LOCATION MATPLOTLIB FULL

The point is that the PR you merged didn't get full approval, and should not have gone in. Until the approval is collective, you cannot force PRs like this (see ).

The debate had to be done in #3855, and we should have found a satisfactory solution there. We running in circles here, and you are reverting the situation to your advantage. We also loose a lot of usable space with blank in the legend box. On the left bar-plot, the legend makes the bar look like an horizontal bar, I find it more confusing than the previous legend symbol which was more square-ish. You could say it's bad cause there is overlapping with data, but I say it's very consistent with default plotting I (and others) think that the previous, even if not perfect, was better. Here is a complete example to show the usage import pandas as pdĭf.It is your sole opinion, and I disagree. Even when using - there is no parameter which adjusts legend position, only if the legend is drawn. Therefore, the only way to do this at present is to use plt.legend() directly - as outlined in my original answer, below.Īs the comments indicate, you have to use plt.legend(loc='lower left') to put the legend in the lower left. None of those methods allow for legend positioning via their keyword arguments. The remaining **kwargs are passed into the underlying matplotlib.pyplot method which corresponds to the specified 'kind' argument (defaults to ).

It does not accept legend position strings. In its current implementation (version 1.2.3) the 'legend' argument of plot accepts only a boolean or the string 'reverse': To clarify the original answer, there is presently no way to do this through.

Subplot legend consistent legend location matplotlib