question-mark
Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

Vertical size of Bokeh plots makes performance report illegible

See original GitHub issue

What happened: Vertical size of Bokeh plots (Task Stream, Worker Profile, etc.) is extremely large and plots are therefore not legible. Zooming browser out to 25% still doesn’t get the whole plot in – and at that point text becomes impossible to read. Using Brave on OSX. Chrome is slightly better but not much.

What you expected to happen: Plots (both visuals and text) to be legible, ideally without having to zoom out.

Minimal Complete Verifiable Example:

import dask
from dask.distributed import performance_report, Client

client = Client(n_workers=50)
df = dask.datasets.timeseries('2000', '2001')
df = df.persist()

with performance_report(filename="dask-report.html"):
    df.groupby('name').x.mean().compute()

Anything else we need to know?:

Environment:

  • Dask version: 2021.07.1
  • Python version: 3.8
  • Operating System: OS 11.3.1
  • Install method (conda, pip, source): conda

Screenshot attached Screenshot 2021-07-28 at 17 36 51

Issue Analytics

  • State:closed
  • Created 2 years ago
  • Comments:14 (11 by maintainers)

github_iconTop GitHub Comments

1reaction
jrbourbeaucommented, Jul 29, 2021

There seem to be a couple of conversation threads happening at the same time. @rrpelgrim @ncclementi see https://github.com/dask/distributed/issues/5132#issuecomment-888497092 and https://github.com/dask/distributed/issues/5132#issuecomment-888527515 where we determined the issue is related to the size of the “Scheduler Logs” panel. You can increase the number of workers (e.g. Client(n_workers=50)) which will increase the size of the scheduler logs and result in the vertical size issue appearing

1reaction
jrbourbeaucommented, Jul 28, 2021

yup, using Client(n_workers=100) on main with bokeh 2.3.3 causes the problem.

Thanks @gjoseph92 for confirming the scheduler log tab is the issue

I’d say that with the example above, I noticed that when reducing the window size, there is a moment that it stops scaling the plot and cuts it off, so you would need to scroll.

Good observation. This is probably a good thing as, for very small screens, scrolling seems like a more legible alternative to having super tiny figures. I think what we want is to have the scheduler log tab follow a stretch_both sizing mode where it’s doesn’t result in other panels in the tabs being elongated. cc @charlesbluca for the scheduler logs connection and @bryevdv for visibility

Read more comments on GitHub >

github_iconTop Results From Across the Web

plots — Bokeh 2.4.3 Documentation
The (default) data range of the vertical dimension of the plot. What kind of scale to use to convert x-coordinates in data space...
Read more >
how to adjust # of ticks on Bokeh axis (labels are overlapping ...
I have a multi-figure Bokeh plot of vertically stacked & aligned figures. Because I want to align the plots vertically, the y-axis labels...
Read more >
Make an area plot in Python using Bokeh - GeeksforGeeks
1. varea() method: varea() method is a vertical directed area which has one x coordinate array and two y coordinate arrays, y1 and...
Read more >
Python Guide: Bokeh Cheat Sheet - Pluralsight
Steps to Create a Plot Using Bokeh.plotting · Importing the library and its API · Preparing the data · Specifying how and where...
Read more >
Linked Crosshair - HoloViews - HoloViz Discourse
Can't get CrossHairTool in Bokeh to be linked over several plots ... CrosshairTool(dimensions="both") # make a hook to manipulate bokeh figure properties ...
Read more >

github_iconTop Related Medium Post

No results found

github_iconTop Related StackOverflow Question

No results found

github_iconTroubleshoot Live Code

Lightrun enables developers to add logs, metrics and snapshots to live code - no restarts or redeploys required.
Start Free

github_iconTop Related Reddit Thread

No results found

github_iconTop Related Hackernoon Post

No results found

github_iconTop Related Tweet

No results found

github_iconTop Related Dev.to Post

No results found

github_iconTop Related Hashnode Post

No results found