Merge pull request #217 from easyscience/develop

🎉 Release: merge develop into master

Author: AndrewSazonov

.github/workflows/pypi-test.yml

@@ -40,29 +40,43 @@ jobs:
       - name: Init pixi project
         run: pixi init easydiffraction
 
-      - name: Set the minimum system requirements
+      - name: Configure Pixi platforms
         working-directory: easydiffraction
+        shell: bash
         run: |
-          pixi project system-requirements add macos 14.0
-          pixi project system-requirements add glibc 2.35
-
-      - name: Add Python 3.14 from Conda
+          set -euo pipefail
+
+          case "$RUNNER_OS" in
+            Linux)
+              pixi_platforms='[{ name = "linux-64-glibc-2-35", platform = "linux-64", glibc = "2.35" }]'
+              ;;
+            macOS)
+              pixi_platforms='[{ name = "osx-arm64-macos-14-0", platform = "osx-arm64", macos = "14.0" }]'
+              ;;
+            Windows)
+              pixi_platforms='["win-64"]'
+              ;;
+            *)
+              echo "Unsupported runner OS: $RUNNER_OS"
+              exit 1
+              ;;
+          esac
+
+          PIXI_PLATFORMS="$pixi_platforms" python -c 'import os, pathlib; path = pathlib.Path("pixi.toml"); text = path.read_text(); path.write_text(text.replace(text.split("platforms = ", 1)[1].split("\n", 1)[0], os.environ["PIXI_PLATFORMS"], 1))'
+          pixi workspace platform list
+
+      - name: Add Python 3.14
         working-directory: easydiffraction
         run: pixi add "python=3.14"
 
+      # GNU Scientific Library (required by diffpy.pdffit2)
+      # libc++ for macOS (required by diffpy.pdffit2)
       - name: Add other Conda dependencies
         working-directory: easydiffraction
         run: |
           pixi add gsl
           pixi add --platform osx-arm64 libcxx
 
-      #- name: Add pycrysfml calculator from custom PyPI index
-      #  working-directory: easydiffraction
-      #  run: |
-      #    echo '' >> pixi.toml
-      #    echo '[pypi-dependencies]' >> pixi.toml
-      #    echo 'pycrysfml = { version = ">=0.2.1", index = "https://easyscience.github.io/pypi/" }' >> pixi.toml
-
       - name: Add easydiffraction (with dev dependencies) from PyPI
         working-directory: easydiffraction
         run: pixi add --pypi "easydiffraction[dev]"

.github/workflows/test.yml

@@ -231,14 +231,32 @@ jobs:
             pixi init easydiffraction_py$py_ver
             cd easydiffraction_py$py_ver
 
-            echo "Setting the minimum system requirements"
-            pixi project system-requirements add macos 14.0
-            pixi project system-requirements add glibc 2.35
-
-            echo "Adding Python $py_ver"
+            echo "Configure Pixi platforms"
+            case "$RUNNER_OS" in
+              Linux)
+                pixi_platforms='[{ name = "linux-64-glibc-2-35", platform = "linux-64", glibc = "2.35" }]'
+                ;;
+              macOS)
+                pixi_platforms='[{ name = "osx-arm64-macos-14-0", platform = "osx-arm64", macos = "14.0" }]'
+                ;;
+              Windows)
+                pixi_platforms='["win-64"]'
+                ;;
+              *)
+                echo "Unsupported runner OS: $RUNNER_OS"
+                exit 1
+                ;;
+            esac
+
+            PIXI_PLATFORMS="$pixi_platforms" python -c 'import os, pathlib; path = pathlib.Path("pixi.toml"); text = path.read_text(); path.write_text(text.replace(text.split("platforms = ", 1)[1].split("\n", 1)[0], os.environ["PIXI_PLATFORMS"], 1))'
+            pixi workspace platform list
+
+            echo "Add Python $py_ver"
             pixi add "python=$py_ver"
 
-            echo "Adding GNU Scientific Library (required by diffpy.pdffit2)"
+            # Add other Conda dependencies
+            
+            echo "Add GNU Scientific Library (required by diffpy.pdffit2)"
             pixi add gsl
 
             # diffpy.pdffit2 wheel links @rpath/libc++.1.dylib, which must be
@@ -248,14 +266,6 @@ jobs:
             echo "Adding libc++ for macOS (required by diffpy.pdffit2)"
             pixi add --platform osx-arm64 libcxx
 
-            # Doing this in a hacky way, as pixi does not support adding 
-            # dependencies from a custom PyPI index with a CLI command without
-            # specifying full wheel name.
-            #echo "Adding pycrysfml from custom PyPI index"
-            #echo '' >> pixi.toml
-            #echo '[pypi-dependencies]' >> pixi.toml
-            #echo 'pycrysfml = { version = ">=0.2.1", index = "https://easyscience.github.io/pypi/" }' >> pixi.toml
-
             echo "Looking for wheel in ../dist/py$py_ver/"
             ls -l "../dist/py$py_ver/"
 
@@ -267,7 +277,7 @@ jobs:
 
             whl_abs_path="$(python -c 'import os,sys; print(os.path.abspath(sys.argv[1]))' "${whl_path[0]}")"
 
-            echo "Adding easydiffraction from: $whl_abs_path"
+            echo "Add easydiffraction from: $whl_abs_path"
             pixi add --pypi "easydiffraction[dev] @ ${whl_abs_path}"
 
             echo "Exiting pixi project directory"

CONTRIBUTING.md

@@ -42,8 +42,7 @@ Please make sure you follow the EasyScience organization-wide
 If you are not planning to contribute code, you may want to:
 
 - 🐞 Report a bug — see [Reporting Issues](#11-reporting-issues)
-- 🛡 Report a security issue — see
-  [Security Issues](#12-security-issues)
+- 🛡 Report a security issue — see [Security Issues](#12-security-issues)
 - 💬 Ask a question or start a discussion at
   [Project Discussions](https://github.com/easyscience/diffraction-lib/discussions)
 

docs/dev/adrs/accepted/background-auto-estimate.md

@@ -368,8 +368,9 @@ one-line summary (§1). Empirical calibration was carried out in Phase 2:
 
 - The Stage-2 tolerance multiplier (`c · σ`, `c = 2`) and the width
   percentile (~75th) are first-cut constants; they were validated — not
-  exhaustively swept — against the representative CWL (`ed-2`) and TOF
-  (`ed-13`) datasets plus the analytic unit cases, and produce sensible
+  exhaustively swept — against the representative CWL
+  (`refine-lbco-hrpt-from-data`) and TOF (`fitting-exercise-si-lbco`)
+  datasets plus the analytic unit cases, and produce sensible
   backgrounds there. Re-tuning stays possible if a future dataset needs
   it.
 - The single Stage-1 method (`arpls`) holds for both validated beam
@@ -485,15 +486,16 @@ line-segment points are a real-world reference: stripping them and
 re-running `auto_estimate()` should reproduce a comparable background
 curve. **Phase 2 outcome:** the functional regression validates two
 representative datasets — CWL
-[`ed-2.py`](../../../../docs/docs/tutorials/ed-2.py) and TOF
-[`ed-13.py`](../../../../docs/docs/tutorials/ed-13.py) — comparing the
-estimated curve against the hand-placed reference to within a fraction
-of the measured signal scale; the single `arpls` default and the
-first-cut constants hold for both. Sloping and curved backgrounds are
-covered against exact analytic ground truth by the unit tests, not the
-corpus. A broader per-tutorial sweep (e.g. `ed-17`, `ed-16`) was not
-needed and stays available if a future dataset misbehaves. These checks
-run at the functional / unit level.
+[`refine-lbco-hrpt-from-data.py`](../../../../docs/docs/tutorials/refine-lbco-hrpt-from-data.py)
+and TOF
+[`fitting-exercise-si-lbco.py`](../../../../docs/docs/tutorials/fitting-exercise-si-lbco.py)
+— comparing the estimated curve against the hand-placed reference to
+within a fraction of the measured signal scale; the single `arpls`
+default and the first-cut constants hold for both. Sloping and curved
+backgrounds are covered against exact analytic ground truth by the unit
+tests, not the corpus. A broader per-tutorial sweep (e.g. `ed-17`,
+`ed-16`) was not needed and stays available if a future dataset
+misbehaves. These checks run at the functional / unit level.
 
 The estimator module mirrors into
 `tests/unit/easydiffraction/datablocks/experiment/categories/background/`

docs/dev/adrs/accepted/iucr-cif-tag-alignment.md

@@ -276,8 +276,10 @@ In `analysis/analysis.cif`:
 - Bayesian diagnostics, success/message/iterations/fitting*time,
   `result_kind`, `point_estimate_name`, fit-parameter posterior
   summaries, and the `_alias` / `_constraint` / `_joint_fit` /
-  `\_sequential_fit*`registries — **stay under their current category names**. File-scoping to`analysis/analysis.cif`carries the disambiguation; no`\_easydiffraction\*\*`
-  prefix is added in the default save.
+  `\_sequential_fit*`registries — **stay under their current category
+  names**. File-scoping to`analysis/analysis.cif`carries the
+  disambiguation; no`\_easydiffraction\*\*` prefix is added in the
+  default save.
 - The `_minimizer.*`, `_fitting_mode.*`, `_calculator.*` selectors stay
   under their current names for the same reason.
 

docs/dev/adrs/accepted/project-summary-rendering.md

@@ -1130,7 +1130,7 @@ Summary, Software, Refinement, Structures (one subsection per phase),
 Experiments (one subsection per experiment). Tables use `booktabs`
 (`\toprule`/`\midrule`/`\bottomrule`) and `float`'s `[H]` placement;
 math uses inline LaTeX (`$Fd\bar{3}m$`, `\AA`, `$\deg$`). Reference
-example at [`tmp/latex/example.tex`](../../../../tmp/latex/example.tex).
+example at `tmp/latex/example.tex`.
 
 The `iucrjournals.cls` choice has two practical advantages over a bare
 `article`:

docs/dev/adrs/accepted/switchable-category-owned-selectors.md

@@ -164,7 +164,7 @@ single-purpose blocks aligned with the new `_<cat>.type` rule.
 
 `_minimizer.optimizer_name` and `_minimizer.method_name` are also
 **dropped**. Inspecting
-[`src/easydiffraction/analysis/minimizers/lmfit_leastsq.py`](../../../src/easydiffraction/analysis/minimizers/lmfit_leastsq.py)
+[`src/easydiffraction/analysis/minimizers/lmfit_leastsq.py`](../../../../src/easydiffraction/analysis/minimizers/lmfit_leastsq.py)
 (and the matching `bumps_lm.py`, `dfols.py`, …) shows that `name`
 defaults to the enum tag itself and `method` to a per-engine
 module-level constant. The public API never overrides them at
@@ -370,7 +370,7 @@ class PeakBase(CategoryItem, SwitchableCategoryBase):
 Categories that want to show aliases alongside canonical tags (peak)
 override `show_supported()` to add a third column; the per-category
 override pattern is the same as the existing
-[`base.show_peak_profile_types()`](../../../src/easydiffraction/datablocks/experiment/item/base.py)
+[`base.show_peak_profile_types()`](../../../../src/easydiffraction/datablocks/experiment/item/base.py)
 implementation, just moved onto the category.
 
 `type` is backed by a **real `StringDescriptor`** named `_type` that
@@ -382,12 +382,12 @@ the staleness checks.
 
 For `CategoryItem` substrates (minimizer, peak, extinction, calculator,
 chart, table, fitting_mode) the generic CIF emit/read path
-[`io/cif/serialize.py:170`](../../../src/easydiffraction/io/cif/serialize.py)
+[`io/cif/serialize.py:170`](../../../../src/easydiffraction/io/cif/serialize.py)
 picks the descriptor up by name automatically — no custom hook is
 needed. For the `CategoryCollection` substrate (background only),
-[`category.py:230`](../../../src/easydiffraction/core/category.py)'s
+[`category.py:230`](../../../../src/easydiffraction/core/category.py)'s
 `parameters` returns only loop-item parameters and
-[`io/cif/serialize.py:244`](../../../src/easydiffraction/io/cif/serialize.py)
+[`io/cif/serialize.py:244`](../../../../src/easydiffraction/io/cif/serialize.py)
 writes only the loop, so a collection-level `_type` descriptor needs a
 small additional path: the writer emits the scalar tag above the loop,
 and the reader peeks the scalar before iterating items. This is a
@@ -451,8 +451,8 @@ class ExperimentBase:
 ```
 
 The existing owner-level `show_<cat>_types()` methods
-([`bragg_pd.show_background_types()`](../../../src/easydiffraction/datablocks/experiment/item/bragg_pd.py),
-[`base.show_peak_profile_types()`](../../../src/easydiffraction/datablocks/experiment/item/base.py),
+([`bragg_pd.show_background_types()`](../../../../src/easydiffraction/datablocks/experiment/item/bragg_pd.py),
+[`base.show_peak_profile_types()`](../../../../src/easydiffraction/datablocks/experiment/item/base.py),
 `Calculation.show_calculator_types()`,
 `Analysis.show_supported_minimizer_types()`, …) are **deleted**. The
 mixin's `show_supported()` reproduces the same `['*', tag, description]`
@@ -722,7 +722,7 @@ Notes on the in-scope rows:
   renamed `calculator_type` → `type`, and the CIF block changes from
   `_calculation.*` to `_calculator.*`. The setter delegation pattern is
   already in place today
-  ([`calculation/default.py:61`](../../../src/easydiffraction/datablocks/experiment/categories/calculation/default.py)),
+  ([`calculator/default.py:61`](../../../../src/easydiffraction/datablocks/experiment/categories/calculator/default.py)),
   so no mechanism change is required.
 - Rows 6 and 7 involve §8a's `Rendering` → `Chart` + `Table` split
   (Python category restructure, CIF block split).
@@ -958,7 +958,7 @@ Three things change in this block:
 - `_peak.profile_type` becomes `_peak.type`; the existing
   `_peak.broad_gauss_*` and `_peak.broad_lorentz_*` parameter tags are
   unchanged (the names come from
-  [`src/easydiffraction/datablocks/experiment/categories/peak/cwl_mixins.py`](../../../src/easydiffraction/datablocks/experiment/categories/peak/cwl_mixins.py)).
+  [`src/easydiffraction/datablocks/experiment/categories/peak/cwl_mixins.py`](../../../../src/easydiffraction/datablocks/experiment/categories/peak/cwl_mixins.py)).
   The CIF value is the **canonical tag** (`cwl-pseudo-voigt` here, since
   the example experiment is constant-wavelength); the writable Python
   setter `experiment.peak.type` accepts the alias `'pseudo-voigt'` too

docs/dev/adrs/suggestions/cif-numeric-precision.md

@@ -8,10 +8,10 @@ Core model.
 
 > This ADR follows [`AGENTS.md`](../../../../AGENTS.md). It is the
 > data-side counterpart to
-> [`plotting-docs-performance.md`](plotting-docs-performance.md), which
-> handles **display** precision (downcasting plot arrays to float32).
-> This ADR concerns the precision of numbers we **store and serialize**
-> in CIF, which is a separate decision because CIF is a data
+> [`plotting-docs-performance.md`](../accepted/plotting-docs-performance.md),
+> which handles **display** precision (downcasting plot arrays to
+> float32). This ADR concerns the precision of numbers we **store and
+> serialize** in CIF, which is a separate decision because CIF is a data
 > source-of-truth, not a throwaway view.
 
 ## Context

docs/dev/issues/closed/add-help-methods-to-public-discovery-facades.md

@@ -7,4 +7,4 @@ that do not inherit the guarded object hierarchy: `project.display`,
 Introduced `render_object_help()` so these helpers share the same
 property and method table style as `GuardedBase.help()`. Documented the
 convention in
-[`help-discoverability.md`](../adrs/accepted/help-discoverability.md).
+[`help-discoverability.md`](../../adrs/accepted/help-discoverability.md).

docs/dev/issues/closed/consistent-type-suffix-in-switchable-category-api-names.md

@@ -1,7 +1,7 @@
 # 76. Consistent `_type` Suffix in Switchable-Category API Names
 
 Closed by
-[`switchable-category-owned-selectors.md`](../adrs/accepted/switchable-category-owned-selectors.md).
+[`switchable-category-owned-selectors.md`](../../adrs/accepted/switchable-category-owned-selectors.md).
 The public suffix is dropped: owners expose categories, categories
 expose a uniform `type` property plus `show_supported()`, and CIF uses
 one `_<cat>.type` selector tag per category.