Add GNSS velocity-field analysis toolbox - #46
Merged
Merged
Conversation
Standalone analysis modules for GNSS velocity fields, operating on plain numpy/pandas so they compose with the existing gps_sources downloaders: - trend: MLE trend estimation under power-law + white noise (Bos 2013, Williams 2003), with annual/semi-annual terms and step handling - collocation: least-squares collocation for velocity interpolation - euler: Euler-pole / rigid-plate rotation fitting - gps_imaging: median spatial-structure-function interpolation (GPS imaging) - strain: strain-rate tensor from a velocity field - variability: spatial structure functions and semivariograms - cme: common-mode error estimation - steps / synthetic: step handling and synthetic series generators - validation: GNSS-vs-InSAR comparison, scatter and semivariogram plots - supporting additions to analysis, workflows, io, plotting, processing, quality, rates Docs: analysis-modules.md plus geepers_tour and gnss_insar_validation notebooks, wired into the mkdocs nav. scipy added as a runtime dependency (already present transitively in pixi.lock). Notebooks excluded from ruff (rendered documentation). Tests: 209 passed, 5 skipped. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
|
Check out this pull request on See visual diffs & provide feedback on Jupyter Notebooks. Powered by ReviewNB |
setup-pixi runs `pixi install --locked`, so the lock must match the manifest. Regenerated with pixi 0.48.2 to keep the v6 format (0.72 would bump to v7, which the CI-pinned 0.48.2 cannot read). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The suite runs with filterwarnings=error; CI's locked pandas/pyproj emit warnings that newer local versions don't: - steps.detect_steps_enu: skip empty per-component frames before concat (pandas deprecates concatenating empty/all-NA entries) - pyproj 3.7 raises a numpy "ndim > 0 to scalar" DeprecationWarning on single-point Geod/Transformer calls (collocation, euler); add a filterwarnings ignore scoped to the pyproj module so it cannot mask array-to-scalar bugs in our own code Verified in the CI-matched pixi test env: 209 passed, 5 skipped. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
mgovorcin
added a commit
that referenced
this pull request
Jul 13, 2026
Bring the analysis-toolbox documentation authored alongside the code but not included in #46: - how-to-guides.md: validate vs GPS, network screening, velocity uncertainty, interpolation, plate-boundary-constrained interpolation (resolves the dead anchor links from analysis-modules.md) - background-theory.md: theory for the trend / collocation / euler methods - tutorials.md, getting-started.md, index.md: toolbox coverage + polish - fix a malformed mkdocstrings link in getting-started.md All pages already exist in the mkdocs nav, so this is a content update. mkdocs build is clean (no unresolved links). Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
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.
Summary
Lands the standalone GNSS velocity-field analysis toolbox as a clean PR off
main(post-#45), separated from the old viewer WIP that shared theunr-grid-viewer-updatesbranch.New modules operate on plain numpy/pandas so they compose with the existing
gps_sourcesdownloaders:trendcollocationeulergps_imagingstrainvariabilitycmesteps/syntheticvalidationPlus supporting additions to
analysis,workflows,io,plotting,processing,quality,rates.Docs
docs/analysis-modules.md+geepers_tourandgnss_insar_validationnotebooks, wired into the mkdocsnav.Dependencies
trend,collocation,variability,gps_imaging). It is already resolved inpixi.locktransitively, so the lock is unchanged — avoiding a v6→v7 format bump incompatible with the CI-pinned pixiv0.48.2. CI's non-lockedpixi installreconciles the manifest spec.Validation
black+ruff(pinned v0.11.13) + all pre-commit hooks pass.A,C,Css, …) allowed via per-fileN803ignores, matching the repo's existingN806"for math" convention.Reconstruction note
unr-grid-viewer-updatesbranched before #45, so it carried the stale pre-#45 viewer. This PR was rebuilt fresh on currentmain, bringing only the toolbox code/tests/docs and leaving #45'sgps_sourcesgrid data layer untouched (verified the toolbox has nogps_sourcescoupling).🤖 Generated with Claude Code