Skip to content

Conversation

@ffreyer
Copy link
Collaborator

@ffreyer ffreyer commented Apr 5, 2025

Description

Alternative to #4849. I'm not confident this is a good/working solution so I'm separating it from the pr.

Type of change

Delete options that do not apply:

  • Bug fix (non-breaking change which fixes an issue)
  • New feature (non-breaking change which adds functionality)
  • Breaking change (fix or feature that would cause existing functionality to not work as expected)

Checklist

  • Added an entry in CHANGELOG.md (for new features and breaking changes)
  • Added or changed relevant sections in the documentation
  • Added unit tests for new algorithms, conversion methods, etc.
  • Added reference image tests for new plotting functions, recipes, visual options, etc.
  • test Float64 compat

@github-project-automation github-project-automation bot moved this to Work in progress in PR review Apr 5, 2025
@MakieBot
Copy link
Collaborator

MakieBot commented Apr 5, 2025

Benchmark Results

SHA: 6cd5e7e3507fafb399ae4357c84b69b6afced481

Warning

These results are subject to substantial noise because GitHub's CI runs on shared machines that are not ideally suited for benchmarking.

GLMakie
CairoMakie
WGLMakie

@ffreyer
Copy link
Collaborator Author

ffreyer commented Apr 5, 2025

In W/GLMakie I can zoom out until Axis ticks break with this (which happens around 2^-1000)
image

@ffreyer
Copy link
Collaborator Author

ffreyer commented Apr 9, 2025

This currently has two problems:

  • GLMakie CI is more fussy about the coordinates than I expected. Locally limiting values to a few magnitudes less than floatmax(Float32) works fine but in CI it does not. I don't know why that is and it worries me for compatibility
  • Limiting relative to floatmax(Float32) makes sense for what gets pushed on the GPU but it does not make sense with Float64 values/float32convert. It should be possible to make a barplot from -1e100 to 1e100. Just bumping the limit to ~floatmax(Float64) will break plots that are not in a Float64 range, and applying float32convert manually doesn't work because you can't skip it later

@Moelf
Copy link
Contributor

Moelf commented Jun 30, 2025

do you intend to finish this 👀 ?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

Status: Work in progress

Development

Successfully merging this pull request may close these issues.

4 participants