When using %matplotlib notebook in Jupyter Notebooks on a HiDPI screen (tested on multiple MacBooks of different generations), the plot stays half-sized until the cell is finished with all its computations.
Code for reproduction
%matplotlibnotebookimportmatplotlib.pyplotaspltimportnumpyasnpimporttimedefdraw(ax):
"""Scatter some new data into an axis."""ax.cla()
ax.scatter(*np.random.rand(2, 1000))
ax.get_figure().canvas.draw()
_, ax=plt.subplots(1, 1)
for_inrange(10):
# Draw some new datadraw(ax)
# Do some intermediate computationstime.sleep(.1)
# Plot wont upscale till this sleep is overtime.sleep(10)
Actual outcome
Plots stay small until last sleep finishes.
Expected outcome
Plots should rescale to screen's DPI right after initial drawing.
Matplotlib version
Operating system: Ubuntu 16.04 / macOS 10.13
Matplotlib version: 2.1.2
Matplotlib backend: nbAgg
Python version: 3.6.4
Jupyter version: 4.4.0
Other Information
#5383 is probably relevant for when attempting to fix this problem.
There is something not quite right with the way we handle the negotiation about what the screen DPI is and do not fix it until the second draw. The way that draw gets triggered as the js side sends a message back asking for a re-draw, but we do not process that request until your code stops blocking.
The first issue can probably be fixed, but the second probably can not (see #4779 for longer discussion) be (at least in general and not without work on the jupyter side as well.
Would you be interested in working on fixing this?
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!
Activity
tacaswell commented on Feb 14, 2018
There is something not quite right with the way we handle the negotiation about what the screen DPI is and do not fix it until the second draw. The way that draw gets triggered as the js side sends a message back asking for a re-draw, but we do not process that request until your code stops blocking.
The first issue can probably be fixed, but the second probably can not (see #4779 for longer discussion) be (at least in general and not without work on the jupyter side as well.
Would you be interested in working on fixing this?
Also have a look at https://github.com/matplotlib/jupyter-matplotlib which is built on top of ipywidgets and may behave better.
github-actions commented on May 1, 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!