This issue has been marked "inactive" because it has been 365 days since the last comment. If this issue is still present in recent Matplotlib releases, or the feature request is still wanted, please leave a comment and this label will be removed. If there are no updates in another 30 days, this issue will be automatically closed, but you are free to re-open or create a new issue if needed. We value issue reports, and this procedure is meant to help us resurface and prioritize issues that have not been addressed yet, not make them disappear. Thanks for your help!
Is this still really necessary given the intervening 10 years and much better computers? I guess for some visualizations this could be dealt with using the data pipeline and adaptive resampling of the underlying data.
To improve interactive response during pan/zoom, NOT updating the axes will give better result (by using blit method).
I use both approach (reduce quality + no axes update) in my github project and the axes update is the most consuming.
Axis updates are indeed expensive. We have a long-term idea how to improve on this, but it requires significant rearchitecting.
Tip: For better user feedback, you could gray-out the ticks during panning, so that users immediately see that the ticks are not valid. This could be done by changing the tick color at the pan start/end, which would require one additional draw at the start. Or, since you have an overlay already, you could maybe overlay all the ticks with a semi-transparent rectangle with the figure background color.
Activity
petehuang commented on Dec 31, 2016
Assuming this is still fair game despite how old the issue is. I wonder about the priority though.
github-actions commented on Mar 2, 2023
This issue has been marked "inactive" because it has been 365 days since the last comment. If this issue is still present in recent Matplotlib releases, or the feature request is still wanted, please leave a comment and this label will be removed. If there are no updates in another 30 days, this issue will be automatically closed, but you are free to re-open or create a new issue if needed. We value issue reports, and this procedure is meant to help us resurface and prioritize issues that have not been addressed yet, not make them disappear. Thanks for your help!
tacaswell commented on Mar 2, 2023
This is still an interesting idea and still a lot of work!
If we expect more actually remote ipympl usage this is more important that it used to be due to the inherent latency.
The tricky thing will be to degrade the artists in the Axes without degrading the tick labels.
jklymak commented on Mar 2, 2023
Is this still really necessary given the intervening 10 years and much better computers? I guess for some visualizations this could be dealt with using the data pipeline and adaptive resampling of the underlying data.
tacaswell commented on Mar 2, 2023
If you give a scientist a better computer they find bigger data to fill it :-p
We have also "lost" a factor of 4 in rendering cost due to the rise of hi-dpi screens.
jklymak commented on Mar 3, 2023
As pointed out on the call by @anntzer perhaps the lowest hanging fruit is the transform stack, which is incredibly flexible, but not super efficient.
mp-007 commented on Mar 10, 2026
To improve interactive response during pan/zoom, NOT updating the axes will give better result (by using blit method).
I use both approach (reduce quality + no axes update) in my github project and the axes update is the most consuming.
Here's an example for the pan behavior
2024-07-10.15-09-57.1.mp4timhoffm commented on Mar 10, 2026
Nice!
Axis updates are indeed expensive. We have a long-term idea how to improve on this, but it requires significant rearchitecting.
Tip: For better user feedback, you could gray-out the ticks during panning, so that users immediately see that the ticks are not valid. This could be done by changing the tick color at the pan start/end, which would require one additional draw at the start. Or, since you have an overlay already, you could maybe overlay all the ticks with a semi-transparent rectangle with the figure background color.
mp-007 commented on Mar 10, 2026
Thanks for the tip! I will add this to my TODO list.
I'm always open to any improvement or features