On 4/17/12 9:29 PM, Dan Drake wrote:
On Tue, 17 Apr 2012 at 08:18PM -0500, Jason Grout wrote:
Can you try commenting out these 2 lines in plot.py:
# tight_layout adjusts the *subplot* parameters so ticks
aren't cut off, etc.
figure.tight_layout()
in the current beta and seein
On Tue, 17 Apr 2012 at 08:18PM -0500, Jason Grout wrote:
> Can you try commenting out these 2 lines in plot.py:
>
> # tight_layout adjusts the *subplot* parameters so ticks
> aren't cut off, etc.
> figure.tight_layout()
>
> in the current beta and seeing if that fixes the problem?
On 4/17/12 8:16 PM, kcrisman wrote:
So we need to look at the prealphas (which I don't have built) and see
what happened between 4.8 and 5.0.beta1.
That certainly makes the Python upgrade or the mpl upgrade the most
likely suspects. I can't believe that
figure.tight_layout()
caused that m
On 4/17/12 8:00 PM, Dan Drake wrote:
On Tue, 17 Apr 2012 at 11:32AM +0200, Jeroen Demeyer wrote:
Now that we're (hopefully) nearing the sage-5.0 release, I am doing
some timings again. We don't yet have proper automatic timing testing,
but with some simple scripting I discovered that plotting ha
> So we need to look at the prealphas (which I don't have built) and see
> what happened between 4.8 and 5.0.beta1.
>
That certainly makes the Python upgrade or the mpl upgrade the most
likely suspects. I can't believe that
figure.tight_layout()
caused that much slowdown, but testing shows tha
I can confirm this with other examples, about a 25-50% slowdown. I'm
sorry I don't have any intermediate versions to bisect this with
currently, as you are right that this is not good.
On Apr 17, 5:52 am, Jeroen Demeyer wrote:
> An easy way to test this using timeit() (best out of 20 runs):
>
>
