Repository navigation
Conversation
…D#674) ubi_to_rgb(ubis, group, view, key, axis) gives TSL and HSV inverse pole figure colours for all 32 point groups, their settings and the R groups on rhombohedral axes. No B matrix or Cartesian frame: the fold uses integer operations on uvw, angles use the symmetrised metric (warns above 5e-3 strain). key_image / plot_key draw the legend. The keys (ImageD11/ipf_keys.py) and the check colours (test/ipf_oracle.json) are generated from MATLAB MTEX 7.2.0 by test/ipf_keys/. Differences from MTEX, measured on the check directions: the hue integral is exact (MTEX: 5.5e-5), the 4-corner centre is the area centroid (TSL m-3, 23: 4.7e-3), and the HSV hue correction starts at 0 (MTEX starts at 1.7e-3 and jumps on one half circle: up to 9.5e-3). Maths: docs/sphx/ipf_colour_math.md. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The sector of a key is the same figure in every metric its group allows (tested to 1e-12 at 20 % strain), so the hue integral is tabulated once per call and each grain only integrates the rest of one piece. Colours change by 2.3e-13; 1e6 ubi take 2-8 s on one core instead of 14-21 s. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
unitcell.get_ipf_colour*, grain.get_ipf_colour, nb_utils.get_rgbs_for_grains, TensorMap.get_ipf_maps and the nb_utils inverse pole figure plots use rgb_colors instead of orix, with key="tsl" or "hsv". The point group comes from the space group (unitcell.ipf_group). Colours from ubi use each ubi's own metric (grains, TensorMap voxels); from U or an orix orientation, the unit cell's. rgb_colors.plot_ipf draws grains on the key, zoomed to the sector. orix orientations stay (get_orix_orien*, get_ubi_from_orix_orien). get_ipf_maps now skips voxels with a NaN UBI (it raised before when any were present). Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Member
Author
|
I suggest we don't merge without some way to refactor and manage the complexity (e.g. move into a sub-module with it's own tests in there) |
ImageD11/rgb_colors/ holds the code, key table, maths doc, tests and the key-generation tools. test/test_rgb_colors.py imports its tests for CI. unitcell, grain, nb_utils and TensorMap take backend="rgb_colors" (default) or "orix" (the previous orix TSL code, unchanged). Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Member
Author
|
@jadball : In theory this should be ready to merge now, and replace the other open IPF PR. But it is Claude written and largely untested. For now, I don't have a clear idea how to recognise correct colors...! |
Contributor
|
I trust it better than code I can write - please feel free to merge. There are some basic sensible checks you can do as a human - pick a unit cell, align one of its real-space axes (a,b,c) to a direction in the sample frame (e.g. put b along y), then look at the IPF plot along that sample direction. You should see, for IPF-y, a dot in [010]. For the colouring, it then depends on the key, which different programs choose differently... |
This branch has not been deployed
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Entirely by Claude (deepseek did not succeed).
While impressive, my feeling is that there is currently too much in here to merge as is. The reason to generate it are:
I wonder if this should re-factor as a folder submodule (like ImageD11.rgb_color) with the code bomb contained inside a folder and a user selectable 'backend' for using pymtex vs orix and just the simplified API. Adding so many lines of generated code is a bit of a concern, especially as this essentially duplicates other packages.