feat: version-agnostic analysis scripts with shared version helpers

- scripts/bicorder_common.py (new): single source of truth for historical
  gradient renames (COLUMN_RENAMES), version detection (bicorder_version col
  → version col → data/<type>_<version>/ dir convention), and training-CSV
  auto-selection (find_training_csv)
- classify_readings.py: auto-select training run by recorded bicorder version
  (excludes the input itself to avoid circular training), canonicalize old
  column names, version-mismatch warnings
- bicorder_classifier.py / export_model_for_js.py: share renames + dimension
  loader; instructive error when clustering results are missing;
  bicorder_version recorded in exported models
- compare_analyses.py: CLI (reference + comparison CSVs), rename
  canonicalization so runs of different versions align on shared gradients,
  Descriptor dedup (was silently skewing merges); legacy no-arg audit intact
- scripts/univariate_analysis.py (new): per-protocol/per-gradient averages,
  distributions, summary stats; --img publishes the three README summary
  charts to img/
- sync_readings.sh: defer classifier training to auto-matching; gitignore
  analysis/.venv and __pycache__

analysis/scripts/bicorder_classifier.py

@@ -30,34 +30,7 @@ from sklearn.discriminant_analysis import LinearDiscriminantAnalysis
 import json
 from pathlib import Path
 
-# Path to bicorder.json (relative to this script)
-_BICORDER_JSON = Path(__file__).parent.parent.parent / 'bicorder.json'
-
-# Historical column renames: maps old CSV column names → current bicorder.json names.
-# Add an entry here whenever gradient terms are renamed in bicorder.json.
-_COLUMN_RENAMES = {
-    'Design_elite_vs_vernacular': 'Design_formal_vs_vernacular',
-    'Design_institutional_vs_vernacular': 'Design_formal_vs_vernacular',
-    'Entanglement_exclusive_vs_non-exclusive': 'Entanglement_monopolistic_vs_pluralistic',
-    'Experience_sufficient_vs_insufficient': 'Experience_sufficient_vs_limited',
-    'Experience_Kafka_vs_Whitehead': 'Experience_restraining_vs_liberating',
-}
-
-
-def _load_bicorder_dimensions(bicorder_path=_BICORDER_JSON):
-    """Read DIMENSIONS and KEY_DIMENSIONS from bicorder.json."""
-    with open(bicorder_path) as f:
-        data = json.load(f)
-    dimensions = []
-    key_dimensions = []
-    for category in data['diagnostic']:
-        set_name = category['set_name']
-        for gradient in category['gradients']:
-            dim_name = f"{set_name}_{gradient['term_left']}_vs_{gradient['term_right']}"
-            dimensions.append(dim_name)
-            if gradient.get('shortform', False):
-                key_dimensions.append(dim_name)
-    return dimensions, key_dimensions
+from bicorder_common import COLUMN_RENAMES as _COLUMN_RENAMES, find_training_csv, load_bicorder_dimensions
 
 
 class BicorderClassifier:
@@ -71,19 +44,32 @@ class BicorderClassifier:
         2: "Institutional/Bureaucratic"
     }
 
-    def __init__(self, diagnostic_csv='data/synthetic_1.2.6/readings.csv',
+    def __init__(self, diagnostic_csv=None,
                  model_path=None):
-        """Initialize classifier with pre-computed model data."""
+        """Initialize classifier with pre-computed model data.
+
+        If diagnostic_csv is None, the most recent synthetic readings.csv is
+        selected automatically (see bicorder_common.find_training_csv).
+        """
+        if diagnostic_csv is None:
+            diagnostic_csv = find_training_csv()
+            if diagnostic_csv is None:
+                raise FileNotFoundError(
+                    "No training CSV found under data/*/readings.csv — "
+                    "pass diagnostic_csv explicitly."
+                )
+            print(f"No training CSV specified; using {diagnostic_csv}")
         if model_path is None:
             model_path = str(Path(diagnostic_csv).parent / 'analysis' / 'data')
-        self._diagnostic_csv = diagnostic_csv
+        self._diagnostic_csv = str(diagnostic_csv)
         self.model_path = Path(model_path)
         self.scaler = StandardScaler()
         self.lda = None
         self.cluster_centroids = None
 
-        # Derive dimension lists from bicorder.json
-        self.DIMENSIONS, self.KEY_DIMENSIONS = _load_bicorder_dimensions()
+        # Derive dimension lists (and version) from bicorder.json
+        self.DIMENSIONS, self.KEY_DIMENSIONS, self.bicorder_version = (
+            load_bicorder_dimensions())
 
         # Load training data to fit scaler and LDA
         self._load_model()
@@ -92,7 +78,13 @@ class BicorderClassifier:
         """Load and fit the classification model from analysis results."""
         # Load the original data and cluster assignments
         df = pd.read_csv(self._diagnostic_csv)
-        clusters = pd.read_csv(self.model_path / 'kmeans_clusters.csv')
+        clusters_path = self.model_path / 'kmeans_clusters.csv'
+        if not clusters_path.exists():
+            raise FileNotFoundError(
+                f"No clustering results for training data: {clusters_path} not found. "
+                f"Run `python3 scripts/multivariate_analysis.py {self._diagnostic_csv}` "
+                f"first to generate cluster assignments.")
+        clusters = pd.read_csv(clusters_path)
 
         # Rename old column names to match current bicorder.json
         df = df.rename(columns=_COLUMN_RENAMES)
@@ -283,6 +275,8 @@ class BicorderClassifier:
     def save_model(self, output_path='bicorder_classifier_model.json'):
         """Save model parameters for use without scikit-learn."""
         model_data = {
+            'bicorder_version': self.bicorder_version,
+            'trained_on': self._diagnostic_csv,
             'dimensions': self.DIMENSIONS,
             'key_dimensions': self.KEY_DIMENSIONS,
             'cluster_names': self.CLUSTER_NAMES,