scitex 2.5.0__py3-none-any.whl → 2.7.3__py3-none-any.whl

scitex/stats/effect_sizes/_epsilon_squared.py

@@ -5,6 +5,7 @@
 # ----------------------------------------
 from __future__ import annotations
 import os
+
 __FILE__ = "./src/scitex/stats/effect_sizes/_epsilon_squared.py"
 __DIR__ = os.path.dirname(__FILE__)
 # ----------------------------------------
@@ -154,13 +155,13 @@ def interpret_epsilon_squared(epsilon2: float) -> str:
     'large'
     """
     if epsilon2 < 0.01:
-        return 'negligible'
+        return "negligible"
     elif epsilon2 < 0.06:
-        return 'small'
+        return "small"
     elif epsilon2 < 0.14:
-        return 'medium'
+        return "medium"
     else:
-        return 'large'
+        return "large"
 
 
 """Main function"""
@@ -231,13 +232,13 @@ def main(args):
 
     # Plot 1: Distribution comparison
     ax = axes[0]
-    ax.hist(group1, bins=20, alpha=0.5, label='Group 1', density=True)
-    ax.hist(group2, bins=20, alpha=0.5, label='Group 2', density=True)
-    ax.hist(group3, bins=20, alpha=0.5, label='Group 3', density=True)
+    ax.hist(group1, bins=20, alpha=0.5, label="Group 1", density=True)
+    ax.hist(group2, bins=20, alpha=0.5, label="Group 2", density=True)
+    ax.hist(group3, bins=20, alpha=0.5, label="Group 3", density=True)
 
-    ax.set_xlabel('Value')
-    ax.set_ylabel('Density')
-    ax.set_title(f'Skewed Distributions (ε² = {eps2:.2f})')
+    ax.set_xlabel("Value")
+    ax.set_ylabel("Density")
+    ax.set_title(f"Skewed Distributions (ε² = {eps2:.2f})")
     ax.legend()
 
     # Plot 2: Comparison of methods
@@ -253,17 +254,17 @@ def main(args):
         eps2_values.append(epsilon_squared([g1, g2]))
         eta2_values.append(eta_squared([g1, g2]))
 
-    ax.plot(scales, eps2_values, 'o-', label='Epsilon-squared (ε²)', linewidth=2)
-    ax.plot(scales, eta2_values, 's-', label='Eta-squared (η²)', linewidth=2)
+    ax.plot(scales, eps2_values, "o-", label="Epsilon-squared (ε²)", linewidth=2)
+    ax.plot(scales, eta2_values, "s-", label="Eta-squared (η²)", linewidth=2)
 
-    ax.set_xlabel('Distribution Scale Difference')
-    ax.set_ylabel('Effect Size')
-    ax.set_title('Non-parametric vs Parametric Effect Size')
+    ax.set_xlabel("Distribution Scale Difference")
+    ax.set_ylabel("Effect Size")
+    ax.set_title("Non-parametric vs Parametric Effect Size")
     ax.legend()
     ax.grid(True, alpha=0.3)
 
     stx.plt.tight_layout()
-    stx.io.save(fig, './epsilon_squared_demo.jpg')
+    stx.io.save(fig, "./epsilon_squared_demo.jpg")
     logger.info("Visualization saved")
 
     return 0
@@ -274,11 +275,7 @@ def parse_args():
     parser = argparse.ArgumentParser(
         description="Demonstrate epsilon-squared effect size calculation"
     )
-    parser.add_argument(
-        '--verbose',
-        action='store_true',
-        help='Enable verbose output'
-    )
+    parser.add_argument("--verbose", action="store_true", help="Enable verbose output")
     return parser.parse_args()
 
 
@@ -309,7 +306,7 @@ def run_main():
     )
 
 
-if __name__ == '__main__':
+if __name__ == "__main__":
     run_main()
 
 # EOF

scitex/stats/effect_sizes/_eta_squared.py

@@ -5,6 +5,7 @@
 # ----------------------------------------
 from __future__ import annotations
 import os
+
 __FILE__ = "./src/scitex/stats/effect_sizes/_eta_squared.py"
 __DIR__ = os.path.dirname(__FILE__)
 # ----------------------------------------
@@ -38,10 +39,7 @@ logger = getLogger(__name__)
 """Functions"""
 
 
-def eta_squared(
-    groups: List[Union[np.ndarray, pd.Series]],
-    ddof: int = 1
-) -> float:
+def eta_squared(groups: List[Union[np.ndarray, pd.Series]], ddof: int = 1) -> float:
     """
     Compute eta-squared (η²) effect size for ANOVA.
 
@@ -160,13 +158,13 @@ def interpret_eta_squared(eta2: float) -> str:
     'large'
     """
     if eta2 < 0.01:
-        return 'negligible'
+        return "negligible"
     elif eta2 < 0.06:
-        return 'small'
+        return "small"
     elif eta2 < 0.14:
-        return 'medium'
+        return "medium"
     else:
-        return 'large'
+        return "large"
 
 
 """Main function"""
@@ -212,7 +210,9 @@ def main(args):
     groups = [np.random.normal(i * 0.3, 1, 25) for i in range(5)]
     eta2_many = eta_squared(groups)
 
-    logger.info(f"Five groups: η² = {eta2_many:.3f} ({interpret_eta_squared(eta2_many)})")
+    logger.info(
+        f"Five groups: η² = {eta2_many:.3f} ({interpret_eta_squared(eta2_many)})"
+    )
 
     # Visualization
     logger.info("\n=== Creating visualization ===")
@@ -225,14 +225,14 @@ def main(args):
     data_viz = [group1, group2, group3]
 
     bp = ax.boxplot(data_viz, positions=positions, widths=0.6, patch_artist=True)
-    for patch in bp['boxes']:
-        patch.set_facecolor('lightblue')
+    for patch in bp["boxes"]:
+        patch.set_facecolor("lightblue")
 
-    ax.set_xlabel('Group')
-    ax.set_ylabel('Value')
-    ax.set_title(f'Three Groups (η² = {eta2:.2f})')
-    ax.set_xticklabels(['Group 1', 'Group 2', 'Group 3'])
-    ax.grid(True, alpha=0.3, axis='y')
+    ax.set_xlabel("Group")
+    ax.set_ylabel("Value")
+    ax.set_title(f"Three Groups (η² = {eta2:.2f})")
+    ax.set_xticklabels(["Group 1", "Group 2", "Group 3"])
+    ax.grid(True, alpha=0.3, axis="y")
 
     # Plot 2: Effect size vs number of groups
     ax = axes[1]
@@ -243,14 +243,14 @@ def main(args):
         groups_test = [np.random.normal(i * 0.4, 1, 30) for i in range(n_groups)]
         eta2_values.append(eta_squared(groups_test))
 
-    ax.plot(n_groups_list, eta2_values, 'o-', linewidth=2, markersize=8)
-    ax.set_xlabel('Number of Groups')
-    ax.set_ylabel('Eta-squared')
-    ax.set_title('Effect Size vs Number of Groups')
+    ax.plot(n_groups_list, eta2_values, "o-", linewidth=2, markersize=8)
+    ax.set_xlabel("Number of Groups")
+    ax.set_ylabel("Eta-squared")
+    ax.set_title("Effect Size vs Number of Groups")
     ax.grid(True, alpha=0.3)
 
     stx.plt.tight_layout()
-    stx.io.save(fig, './eta_squared_demo.jpg')
+    stx.io.save(fig, "./eta_squared_demo.jpg")
     logger.info("Visualization saved")
 
     return 0
@@ -261,11 +261,7 @@ def parse_args():
     parser = argparse.ArgumentParser(
         description="Demonstrate eta-squared effect size calculation"
     )
-    parser.add_argument(
-        '--verbose',
-        action='store_true',
-        help='Enable verbose output'
-    )
+    parser.add_argument("--verbose", action="store_true", help="Enable verbose output")
     return parser.parse_args()
 
 
@@ -296,7 +292,7 @@ def run_main():
     )
 
 
-if __name__ == '__main__':
+if __name__ == "__main__":
     run_main()
 
 # EOF

scitex/stats/effect_sizes/_prob_superiority.py

@@ -5,6 +5,7 @@
 # ----------------------------------------
 from __future__ import annotations
 import os
+
 __FILE__ = "./src/scitex/stats/effect_sizes/_prob_superiority.py"
 __DIR__ = os.path.dirname(__FILE__)
 # ----------------------------------------
@@ -39,8 +40,7 @@ logger = getLogger(__name__)
 
 
 def prob_superiority(
-    x: Union[np.ndarray, pd.Series],
-    y: Union[np.ndarray, pd.Series]
+    x: Union[np.ndarray, pd.Series], y: Union[np.ndarray, pd.Series]
 ) -> float:
     """
     Compute probability of superiority P(X > Y).
@@ -170,13 +170,13 @@ def interpret_prob_superiority(prob: float) -> str:
     distance = abs(prob - 0.5)
 
     if distance < 0.06:
-        return 'negligible'
+        return "negligible"
     elif distance < 0.14:
-        return 'small'
+        return "small"
     elif distance < 0.21:
-        return 'medium'
+        return "medium"
     else:
-        return 'large'
+        return "large"
 
 
 """Main function"""
@@ -196,6 +196,7 @@ def main(args):
     y = np.array([3, 4, 5, 6, 7, 8, 9, 10])
 
     from ._cliffs_delta import cliffs_delta
+
     prob = prob_superiority(x, y)
     delta = cliffs_delta(x, y)
 
@@ -231,20 +232,20 @@ def main(args):
         control = np.random.normal(0, 1, 100)
         probs.append(prob_superiority(treatment, control))
 
-    ax.plot(shifts, probs, 'o-', linewidth=2, markersize=8)
-    ax.axhline(0.5, color='red', linestyle='--', alpha=0.5, label='Chance level')
-    ax.axhline(0.56, color='orange', linestyle='--', alpha=0.5, label='Small effect')
-    ax.axhline(0.64, color='yellow', linestyle='--', alpha=0.5, label='Medium effect')
-    ax.axhline(0.71, color='green', linestyle='--', alpha=0.5, label='Large effect')
+    ax.plot(shifts, probs, "o-", linewidth=2, markersize=8)
+    ax.axhline(0.5, color="red", linestyle="--", alpha=0.5, label="Chance level")
+    ax.axhline(0.56, color="orange", linestyle="--", alpha=0.5, label="Small effect")
+    ax.axhline(0.64, color="yellow", linestyle="--", alpha=0.5, label="Medium effect")
+    ax.axhline(0.71, color="green", linestyle="--", alpha=0.5, label="Large effect")
 
-    ax.set_xlabel('Mean Shift (Cohen\'s d)')
-    ax.set_ylabel('P(Treatment > Control)')
-    ax.set_title('Probability of Superiority vs Effect Size')
+    ax.set_xlabel("Mean Shift (Cohen's d)")
+    ax.set_ylabel("P(Treatment > Control)")
+    ax.set_title("Probability of Superiority vs Effect Size")
     ax.legend()
     ax.grid(True, alpha=0.3)
 
     stx.plt.tight_layout()
-    stx.io.save(fig, './prob_superiority_demo.jpg')
+    stx.io.save(fig, "./prob_superiority_demo.jpg")
     logger.info("Visualization saved")
 
     return 0
@@ -255,11 +256,7 @@ def parse_args():
     parser = argparse.ArgumentParser(
         description="Demonstrate probability of superiority calculation"
     )
-    parser.add_argument(
-        '--verbose',
-        action='store_true',
-        help='Enable verbose output'
-    )
+    parser.add_argument("--verbose", action="store_true", help="Enable verbose output")
     return parser.parse_args()
 
 
@@ -290,7 +287,7 @@ def run_main():
     )
 
 
-if __name__ == '__main__':
+if __name__ == "__main__":
     run_main()
 
 # EOF

scitex/stats/io/__init__.py

@@ -0,0 +1,29 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+# Timestamp: "2025-12-13 (ywatanabe)"
+# File: /home/ywatanabe/proj/scitex-code/src/scitex/stats/io/__init__.py
+
+"""
+I/O operations for scitex.stats - Statistical results bundles (.statsz).
+
+This module handles:
+    - .statsz bundle load/save operations
+    - Statistical comparison metadata
+    - P-value and effect size validation
+"""
+
+from ._bundle import (
+    validate_statsz_spec,
+    load_statsz_bundle,
+    save_statsz_bundle,
+    STATSZ_SCHEMA_SPEC,
+)
+
+__all__ = [
+    "validate_statsz_spec",
+    "load_statsz_bundle",
+    "save_statsz_bundle",
+    "STATSZ_SCHEMA_SPEC",
+]
+
+# EOF

scitex/stats/io/_bundle.py

@@ -0,0 +1,156 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+# Timestamp: "2025-12-13 (ywatanabe)"
+# File: /home/ywatanabe/proj/scitex-code/src/scitex/stats/io/_bundle.py
+
+"""
+SciTeX .statsz Bundle I/O - Statistics-specific bundle operations.
+
+Handles:
+    - Statistical results specification validation
+    - Comparison metadata management
+    - P-value and effect size validation
+"""
+
+import json
+from pathlib import Path
+from typing import Any, Dict, List
+
+__all__ = [
+    "validate_statsz_spec",
+    "load_statsz_bundle",
+    "save_statsz_bundle",
+    "STATSZ_SCHEMA_SPEC",
+]
+
+# Schema specification for .statsz bundles
+STATSZ_SCHEMA_SPEC = {
+    "name": "scitex.stats.stats",
+    "version": "1.0.0",
+    "required_fields": ["schema"],
+    "optional_fields": ["comparisons", "metadata", "descriptive", "test_results"],
+}
+
+
+def validate_statsz_spec(spec: Dict[str, Any]) -> List[str]:
+    """Validate .statsz-specific fields.
+
+    Args:
+        spec: The specification dictionary to validate.
+
+    Returns:
+        List of validation error messages (empty if valid).
+    """
+    errors = []
+
+    if "comparisons" in spec:
+        comparisons = spec["comparisons"]
+        if not isinstance(comparisons, list):
+            errors.append("'comparisons' must be a list")
+        else:
+            for i, comp in enumerate(comparisons):
+                if not isinstance(comp, dict):
+                    errors.append(f"comparisons[{i}] must be a dictionary")
+                    continue
+
+                # Validate p_value if present
+                if "p_value" in comp:
+                    p = comp["p_value"]
+                    if not isinstance(p, (int, float)):
+                        errors.append(f"comparisons[{i}].p_value must be numeric")
+                    elif not (0 <= p <= 1):
+                        errors.append(
+                            f"comparisons[{i}].p_value must be between 0 and 1"
+                        )
+
+                # Validate effect_size if present
+                if "effect_size" in comp:
+                    es = comp["effect_size"]
+                    if isinstance(es, dict):
+                        if "value" in es and not isinstance(es["value"], (int, float)):
+                            errors.append(
+                                f"comparisons[{i}].effect_size.value must be numeric"
+                            )
+                    elif not isinstance(es, (int, float)):
+                        errors.append(
+                            f"comparisons[{i}].effect_size must be numeric or dict"
+                        )
+
+    # Validate test_results if present
+    if "test_results" in spec:
+        test_results = spec["test_results"]
+        if not isinstance(test_results, (dict, list)):
+            errors.append("'test_results' must be a dictionary or list")
+
+    # Validate descriptive if present
+    if "descriptive" in spec:
+        descriptive = spec["descriptive"]
+        if not isinstance(descriptive, dict):
+            errors.append("'descriptive' must be a dictionary")
+
+    return errors
+
+
+def load_statsz_bundle(bundle_dir: Path) -> Dict[str, Any]:
+    """Load .statsz bundle contents from directory.
+
+    Args:
+        bundle_dir: Path to the bundle directory.
+
+    Returns:
+        Dictionary with loaded bundle contents.
+    """
+    result = {}
+
+    # Load specification
+    spec_file = bundle_dir / "stats.json"
+    if spec_file.exists():
+        with open(spec_file, "r") as f:
+            result["spec"] = json.load(f)
+    else:
+        result["spec"] = None
+
+    # Load supplementary data files if present
+    data_file = bundle_dir / "data.csv"
+    if data_file.exists():
+        try:
+            import pandas as pd
+            result["data"] = pd.read_csv(data_file)
+        except ImportError:
+            with open(data_file, "r") as f:
+                result["data"] = f.read()
+
+    return result
+
+
+def save_statsz_bundle(data: Dict[str, Any], dir_path: Path) -> None:
+    """Save .statsz bundle contents to directory.
+
+    Args:
+        data: Bundle data dictionary.
+        dir_path: Path to the bundle directory.
+    """
+    # Save specification
+    spec = data.get("spec", {})
+    spec_file = dir_path / "stats.json"
+    with open(spec_file, "w") as f:
+        json.dump(spec, f, indent=2)
+
+    # Save supplementary data if present
+    if "data" in data:
+        data_file = dir_path / "data.csv"
+        df = data["data"]
+        if hasattr(df, "to_csv"):
+            df.to_csv(data_file, index=False)
+        else:
+            with open(data_file, "w") as f:
+                f.write(str(df))
+
+    # Save summary report if present
+    if "report" in data:
+        report_file = dir_path / "report.md"
+        with open(report_file, "w") as f:
+            f.write(data["report"])
+
+
+# EOF

scitex/stats/posthoc/__init__.py

@@ -13,7 +13,7 @@ from ._games_howell import posthoc_games_howell
 from ._dunnett import posthoc_dunnett
 
 __all__ = [
-    'posthoc_tukey',
-    'posthoc_games_howell',
-    'posthoc_dunnett',
+    "posthoc_tukey",
+    "posthoc_games_howell",
+    "posthoc_dunnett",
 ]