Skip to content

Add the symlog(loglinglog) axis to the binning plugin - #5534

Draft
ikbuibui wants to merge 2 commits into
ComputationalRadiationPhysics:devfrom
ikbuibui:feat/symlog
Draft

ikbuibui wants to merge 2 commits into
ComputationalRadiationPhysics:devfrom
ikbuibui:feat/symlog

Conversation

@ikbuibui

Copy link
Copy Markdown
Contributor

Adds a copysign function to PMAcc math.

Adds a symlog axis to the binning plugin (a Log-Lin-Log axis).

This axis has a symmetric linear region around zero which has a user configurable threshold (which is mapped to the range -1 to 1) and is logarithmic outside, with continuous transitions (but not necessarily smooth) at the threshold boundaries.
In the logarithmic range, the values are mapped with magnitude 1 + ln(abs(val)/linThreshold) with the same sign as val.

NOTE The bins are defined as equidistant in the symlog space. I am not sure if this is the best way to define bins with this axis, maybe independently setting the number of bins in the linear and log regions is better or maybe something else.
I am currently open to suggestions.

Added tests for the symlog axis.

@chillenzer

Copy link
Copy Markdown
Contributor

What's actually the status of this?

@ikbuibui

Copy link
Copy Markdown
Contributor Author

It is working, but the interface to configure such an axis isn't aligned with what physicists expect/isnt expressive enough. It is set up currently with a full range, total number of bins, and the linear threshold. It isnt obvious from this interface how the bins are distributed in the linear and log regions.

@ikbuibui

Copy link
Copy Markdown
Contributor Author

Thanks for reminding me about this. I will work on this the next time I want a light distraction

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

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants