Refactor switchable experiment, analysis, and display categories

docs/architecture/issues_closed.md

@@ -117,9 +117,19 @@ to an incompatible data class and silently discarding loaded data.
 
 Added `SEQUENTIAL = 'sequential'` to `FitModeEnum`. `fit_sequential()`
 now sets `fit_mode.mode = 'sequential'` internally so the mode is
-persisted in CIF. Added `show_supported_fit_mode_types()` (filters by
-experiment count: ≤1 → only `single`; >1 → all three) and
-`show_current_fit_mode_type()` on `Analysis`. If `fit()` is called while
-mode is `'sequential'`, it logs an error directing the user to
-`fit_sequential()`. Promoted `fit_mode` from a pure single-type category
-to one with show methods.
+persisted in CIF. Added `show_fit_mode_types()` (filters by experiment
+count: ≤1 → only `single`; >1 → all three) and `show_fit_mode_types()`
+on `Analysis`. If `fit()` is called while mode is `'sequential'`, it
+logs an error directing the user to `fit_sequential()`. Promoted
+`fit_mode` from a pure single-type category to one with show methods.
+
+---
+
+## Persist Per-Experiment Calculator Selection
+
+Calculator selection now lives in the experiment-side `calculation`
+category and is persisted in experiment CIF via
+`_calculation.calculator_type`. Loading restores the explicit backend
+selection before dependent experiment categories are populated, so saved
+projects keep the exact active calculator configuration instead of
+falling back to auto-resolution.

docs/architecture/issues_open.md

@@ -14,7 +14,7 @@ needed.
 
 **Type:** Fragility
 
-`joint_fit_experiments` is created once when `fit_mode` becomes
+`joint_fit_experiments` is created once when `fit.mode` becomes
 `'joint'`. If experiments are added, removed, or renamed afterwards, the
 weight collection is stale. Joint fitting can fail with missing keys or
 run with incorrect weights.
@@ -180,25 +180,6 @@ minimiser.
 
 ---
 
-## 16. 🟡 Persist Per-Experiment `calculator_type`
-
-**Type:** Completeness
-
-The current architecture moved calculator selection to the experiment
-level via `calculator_type`, but this selection is not written to CIF
-during `save()` / `show_as_cif()`. Reloading or exporting a project
-therefore loses explicit calculator choices and falls back to
-auto-resolution.
-
-**Fix:** serialise `calculator_type` as part of the experiment or
-analysis state, and make sure `load()` restores it. The saved project
-should represent the exact active calculator configuration, not just a
-re-derivable default.
-
-**Depends on:** nothing.
-
----
-
 ## 17. 🟢 Use PDF-Specific CIF Names for Total Scattering
 
 **Type:** Naming
@@ -1191,9 +1172,9 @@ check. Options:
 
 **Type:** API completeness
 
-`show_supported_calculator_types()` exists per-experiment, but there is
-no project/analysis-level method to list all available calculator
-engines. Users exploring the API have no single entry point to see what
+`show_calculator_types()` exists per-experiment, but there is no
+project/analysis-level method to list all available calculator engines.
+Users exploring the API have no single entry point to see what
 calculators are installed.
 
 **Depends on:** nothing.
@@ -1206,8 +1187,8 @@ calculators are installed.
 
 The switchable-category naming convention prescribes `<category>_type`
 (getter/setter) and `show_supported_<category>_types()`. But some names
-deviate: e.g. `show_supported_minimizers()` instead of
-`show_supported_minimizer_types()`, and `current_minimizer` instead of
+deviate: e.g. `show_minimizer_types()` instead of
+`show_supported_minimizer_types()`, and `minimizer_type` instead of
 `minimizer_type`. Audit and align all switchable-category APIs.
 
 **Depends on:** nothing.
@@ -1446,7 +1427,6 @@ operation is possible (e.g. in automated pipelines or tests).
 | 13  | Suppress redundant dirty-flag sets               | 🟢 Low   | Performance      |
 | 14  | Finer-grained change tracking                    | 🟢 Low   | Performance      |
 | 15  | Validate joint-fit weights                       | 🟡 Med   | Correctness      |
-| 16  | Persist per-experiment `calculator_type`         | 🟡 Med   | Completeness     |
 | 17  | Use PDF-specific CIF names                       | 🟢 Low   | Naming           |
 | 18  | Move CIF v2→v1 conversion out of calculator      | 🟢 Low   | Maintainability  |
 | 19  | Debug-mode logging for calculator imports        | 🟢 Low   | Diagnostics      |

docs/architecture/package-structure-full.md

@@ -30,14 +30,14 @@
 │   │   │   │   └── 🏷️ class Constraints
 │   │   │   └── 📄 factory.py
 │   │   │       └── 🏷️ class ConstraintsFactory
-│   │   ├── 📁 fit_mode
+│   │   ├── 📁 fit
 │   │   │   ├── 📄 __init__.py
+│   │   │   ├── 📄 default.py
+│   │   │   │   └── 🏷️ class Fit
 │   │   │   ├── 📄 enums.py
 │   │   │   │   └── 🏷️ class FitModeEnum
-│   │   │   ├── 📄 factory.py
-│   │   │   │   └── 🏷️ class FitModeFactory
-│   │   │   └── 📄 fit_mode.py
-│   │   │       └── 🏷️ class FitMode
+│   │   │   └── 📄 factory.py
+│   │   │       └── 🏷️ class FitFactory
 │   │   ├── 📁 joint_fit_experiments
 │   │   │   ├── 📄 __init__.py
 │   │   │   ├── 📄 default.py
@@ -152,6 +152,12 @@
 │   │   │   │   └── 📄 line_segment.py
 │   │   │   │       ├── 🏷️ class LineSegment
 │   │   │   │       └── 🏷️ class LineSegmentBackground
+│   │   │   ├── 📁 calculation
+│   │   │   │   ├── 📄 __init__.py
+│   │   │   │   ├── 📄 default.py
+│   │   │   │   │   └── 🏷️ class Calculation
+│   │   │   │   └── 📄 factory.py
+│   │   │   │       └── 🏷️ class CalculationFactory
 │   │   │   ├── 📁 data
 │   │   │   │   ├── 📄 __init__.py
 │   │   │   │   ├── 📄 bragg_pd.py
@@ -238,6 +244,7 @@
 │   │   │   │   ├── 📄 factory.py
 │   │   │   │   │   └── 🏷️ class PeakFactory
 │   │   │   │   ├── 📄 tof.py
+│   │   │   │   │   ├── 🏷️ class TofPseudoVoigt
 │   │   │   │   │   ├── 🏷️ class TofJorgensen
 │   │   │   │   │   ├── 🏷️ class TofJorgensenVonDreele
 │   │   │   │   │   └── 🏷️ class TofDoubleJorgensenVonDreele
@@ -360,6 +367,14 @@
 │   ├── 📄 __init__.py
 │   └── 📄 ascii.py
 ├── 📁 project
+│   ├── 📁 categories
+│   │   ├── 📁 display
+│   │   │   ├── 📄 __init__.py
+│   │   │   ├── 📄 default.py
+│   │   │   │   └── 🏷️ class Display
+│   │   │   └── 📄 factory.py
+│   │   │       └── 🏷️ class DisplayFactory
+│   │   └── 📄 __init__.py
 │   ├── 📄 __init__.py
 │   ├── 📄 project.py
 │   │   └── 🏷️ class Project

docs/architecture/package-structure-short.md

@@ -19,11 +19,11 @@
 │   │   │   ├── 📄 __init__.py
 │   │   │   ├── 📄 default.py
 │   │   │   └── 📄 factory.py
-│   │   ├── 📁 fit_mode
+│   │   ├── 📁 fit
 │   │   │   ├── 📄 __init__.py
+│   │   │   ├── 📄 default.py
 │   │   │   ├── 📄 enums.py
-│   │   │   ├── 📄 factory.py
-│   │   │   └── 📄 fit_mode.py
+│   │   │   └── 📄 factory.py
 │   │   ├── 📁 joint_fit_experiments
 │   │   │   ├── 📄 __init__.py
 │   │   │   ├── 📄 default.py
@@ -78,6 +78,10 @@
 │   │   │   │   ├── 📄 enums.py
 │   │   │   │   ├── 📄 factory.py
 │   │   │   │   └── 📄 line_segment.py
+│   │   │   ├── 📁 calculation
+│   │   │   │   ├── 📄 __init__.py
+│   │   │   │   ├── 📄 default.py
+│   │   │   │   └── 📄 factory.py
 │   │   │   ├── 📁 data
 │   │   │   │   ├── 📄 __init__.py
 │   │   │   │   ├── 📄 bragg_pd.py
@@ -187,6 +191,12 @@
 │   ├── 📄 __init__.py
 │   └── 📄 ascii.py
 ├── 📁 project
+│   ├── 📁 categories
+│   │   ├── 📁 display
+│   │   │   ├── 📄 __init__.py
+│   │   │   ├── 📄 default.py
+│   │   │   └── 📄 factory.py
+│   │   └── 📄 __init__.py
 │   ├── 📄 __init__.py
 │   ├── 📄 project.py
 │   └── 📄 project_info.py