Object-local frames and per-field entrance-pupil calibration (WIP)#189
Draft
roytsmart wants to merge 1 commit into
Draft
Object-local frames and per-field entrance-pupil calibration (WIP)#189roytsmart wants to merge 1 commit into
roytsmart wants to merge 1 commit into
Conversation
Rework the coordinate-frame handling of `SequentialSystem` and add machinery to calibrate the entrance-pupil extent per field. Frame handling: - `surfaces_all` now composes `self.transformation` into the non-object surfaces (via `na.transformations.compose`), leaving the object at its own transformation, so the whole trace runs in the object frame and `self.transformation` is applied in exactly one place. - `_shoot_rays`/`_ray_error` solve entirely in the first surface's local frame: the surface transformation is applied per trial inside the residual (so the sag `zfunc` is evaluated where it is valid) and once more before returning, fixing a latent bug where a translated/tilted launch surface had its launch `z` clobbered. - The internal seed is broadcast to the full solution shape, so the "rough" seed (aimed at a powered surface, lacking the target-grid axis) no longer leaves the solve under-determined. - `rayfunction_stops` and the pupil trace express their outputs in the object-local frame, and `rayfunction` inverts the composed sensor transformation. Entrance pupil (WIP): - Add `_calc_rayfunction_pupil` and the `_pupil_min`/`_pupil_max` polynomial fits to capture the field dependence of the entrance pupil, plus `_denormalize_pupil`/`_denormalize_field` and a `field_min`/ `field_max` refactor that fits the field extent versus wavelength. Supporting: - `optika.angles` accepts a 2D vector of direction cosines, reconstructing the `z` component from the unit-length constraint. - Add `docs/stop_finding.rst` documenting the stop-finding strategy. Depends on `na.transformations.compose` (sun-data/named-arrays#193). Still WIP: the public `pupil_min`/`pupil_max` are commented out, the pupil calibration traces the dense grid (should be coarse), and several tests fail (finite-object imaging, vignetting, dxf, a grazing field-max tolerance). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01BWk1x6S9g3s8o674HJUqRQ
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
1 participant
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.
Reworks SequentialSystem coordinate frames (object-local; single
self.transformationfolded into non-object surfaces viana.transformations.compose) and adds WIP per-field entrance-pupil calibration.Frame fixes (working):
surfaces_allcomposition, object-local_shoot_rays/_ray_errorsolve with per-trial transform + broadcast seed, object-framerayfunction_stops/pupil outputs, composed-sensor inversion inrayfunction.WIP: per-field pupil fits (
_calc_rayfunction_pupil,_pupil_min/_pupil_max,_denormalize_pupil),optika.angles2D-cosine input,docs/stop_finding.rst.Depends on sun-data/named-arrays#193 (
na.transformations.compose).Known-incomplete: public
pupil_min/pupil_maxcommented out; pupil calibration traces the dense grid (should be coarse) so the full ESIS design does not yet converge; failing tests: finite-object imaging, vignetting, dxf, grazingfield_maxtol. FURST-style systems converge and image.🤖 Generated with Claude Code
https://claude.ai/code/session_01BWk1x6S9g3s8o674HJUqRQ