scitex 2.14.0__py3-none-any.whl → 2.15.2__py3-none-any.whl

scitex/template/_templates/session_plot.py

@@ -0,0 +1,67 @@
+#!/usr/bin/env python3
+"""Plotting session script template."""
+
+TEMPLATE = {
+    "name": "Plotting Session Script",
+    "description": "@stx.session script optimized for figure generation with auto CSV export",
+    "filename": "plot_script.py",
+    "usage": """
+Usage:
+  python plot_script.py
+  python plot_script.py --help
+
+Note: stx.io.save(fig, "plot.png") automatically exports plot.csv with plotted data
+""",
+    "content": '''#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+# Timestamp: "{timestamp}"
+# File: {filepath}
+
+"""
+{docstring}
+
+Note: stx.io.save(fig, "plot.png") automatically exports plotted data to plot.csv
+"""
+
+import numpy as np
+import scitex as stx
+
+
+@stx.session
+def main(
+    n_points=100,
+    CONFIG=stx.INJECTED,
+    plt=stx.INJECTED,
+    COLORS=stx.INJECTED,
+    logger=stx.INJECTED,
+):
+    """Generate publication-ready figures with automatic data export."""
+    # Generate data
+    x = np.linspace(0, 2 * np.pi, n_points)
+    y = np.sin(x)
+
+    # Create figure using stx.plt (records all plotting calls)
+    fig, ax = plt.subplots()
+
+    # Use plot_line for automatic CSV export
+    ax.plot_line(x, y, color=COLORS.blue, label="sin(x)")
+    ax.set_xyt("X (radians)", "Y", "Sine Wave")
+    ax.legend_()
+
+    # Save figure (auto-exports CSV of plotted data)
+    stx.io.save(fig, "figure.png", symlink_to="./data")
+
+    logger.info(f"Figure saved to: {{CONFIG.SDIR_OUT}}/figure.png")
+    logger.info(f"Data exported to: {{CONFIG.SDIR_OUT}}/figure.csv")
+
+    return 0
+
+
+if __name__ == "__main__":
+    main()
+
+# EOF
+''',
+}
+
+# EOF

scitex/template/_templates/session_stats.py

@@ -0,0 +1,77 @@
+#!/usr/bin/env python3
+"""Statistical analysis session script template."""
+
+TEMPLATE = {
+    "name": "Statistical Analysis Session",
+    "description": "@stx.session script for statistical testing with publication-ready output",
+    "filename": "stats_script.py",
+    "usage": """
+Usage:
+  python stats_script.py
+  python stats_script.py --help
+
+Available Tests:
+  stx.stats.test_ttest_ind()      # Independent t-test
+  stx.stats.test_ttest_paired()   # Paired t-test
+  stx.stats.test_anova()          # One-way ANOVA
+  stx.stats.test_wilcoxon()       # Wilcoxon signed-rank
+  stx.stats.test_mannwhitneyu()   # Mann-Whitney U
+""",
+    "content": '''#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+# Timestamp: "{timestamp}"
+# File: {filepath}
+
+"""
+{docstring}
+
+Available Tests (23 total):
+  stx.stats.test_ttest_ind()      # Independent t-test with effect size
+  stx.stats.test_ttest_paired()   # Paired t-test
+  stx.stats.test_anova()          # One-way ANOVA
+  stx.stats.test_wilcoxon()       # Wilcoxon signed-rank (non-parametric)
+  stx.stats.test_mannwhitneyu()   # Mann-Whitney U (non-parametric)
+
+Output Formats:
+  return_as="dataframe"   # pandas DataFrame
+  return_as="dict"        # Python dictionary
+  return_as="latex"       # LaTeX table
+"""
+
+import numpy as np
+import scitex as stx
+
+
+@stx.session
+def main(
+    n_samples=30,
+    effect_size=0.5,
+    CONFIG=stx.INJECTED,
+    plt=stx.INJECTED,
+    logger=stx.INJECTED,
+):
+    """Run statistical analysis with publication-ready output."""
+    # Generate sample data
+    rng = np.random.default_rng(42)
+    group1 = rng.normal(10, 2, n_samples)
+    group2 = rng.normal(10 + effect_size * 2, 2, n_samples)
+
+    # Run t-test with effect size and CI
+    result = stx.stats.test_ttest_ind(group1, group2, return_as="dataframe")
+    logger.info(f"T-test result:\\n{{result}}")
+
+    # Save results to CONFIG.SDIR_OUT
+    stx.io.save(result, "stats_results.csv")
+    logger.info(f"Results saved to: {{CONFIG.SDIR_OUT}}/stats_results.csv")
+
+    return 0
+
+
+if __name__ == "__main__":
+    main()
+
+# EOF
+''',
+}
+
+# EOF

scitex/template/_templates/stats.py

@@ -0,0 +1,323 @@
+#!/usr/bin/env python3
+# Timestamp: 2026-01-25
+# File: src/scitex/template/_templates/stats.py
+# ----------------------------------------
+
+"""Template for stx.stats statistical analysis module usage."""
+
+TEMPLATE = {
+    "name": "Statistics Module",
+    "description": "stx.stats usage for publication-ready statistical analysis (23 tests)",
+    "filename": "stats_script.py",
+    "priority": 3,
+    "content": '''#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+# Timestamp: {timestamp}
+# File: {filepath}
+
+"""
+stx.stats - Publication-Ready Statistics
+========================================
+
+stx.stats provides:
+- 23 statistical tests with proper formatting
+- Automatic normality checking
+- Effect size calculations (Cohen's d, eta², etc.)
+- APA/Nature-style result formatting
+- Power analysis and sample size calculation
+- Multiple comparison corrections
+
+Usage Patterns
+--------------
+1. Direct API: stx.stats.test_*(data, return_as="dataframe")
+2. CLI: scitex stats recommend --n-groups 3 --outcome continuous
+3. MCP: stats_recommend_tests, stats_run_test, stats_format_results
+
+Output Formats
+--------------
+- return_as="dataframe": Full results as pandas DataFrame
+- return_as="dict": Results as dictionary
+- return_as="latex": Ready-to-use LaTeX string
+- return_as="apa": APA-style formatted string
+"""
+
+import numpy as np
+import scitex as stx
+
+# ============================================================
+# Pattern 1: Two-Group Comparisons
+# ============================================================
+
+def two_group_comparison():
+    """Compare two independent groups."""
+    # Sample data
+    group1 = np.random.normal(10, 2, 50)
+    group2 = np.random.normal(12, 2, 50)
+
+    # Independent t-test (with normality check)
+    result = stx.stats.test_ttest_ind(
+        group1, group2,
+        return_as="dataframe"  # Options: dataframe, dict, latex, apa
+    )
+    print("Independent t-test:")
+    print(result)
+    # Returns: t, p, Cohen's d, CI, normality test, power
+
+    # Mann-Whitney U (non-parametric alternative)
+    result_mwu = stx.stats.test_mannwhitneyu(
+        group1, group2,
+        return_as="dataframe"
+    )
+    print("\\nMann-Whitney U:")
+    print(result_mwu)
+
+    # Paired t-test (for matched samples)
+    pre = np.random.normal(10, 2, 30)
+    post = pre + np.random.normal(2, 1, 30)  # Improvement
+
+    result_paired = stx.stats.test_ttest_rel(
+        pre, post,
+        return_as="latex"  # Get LaTeX-formatted output
+    )
+    print("\\nPaired t-test (LaTeX):")
+    print(result_paired)
+
+# ============================================================
+# Pattern 2: Multi-Group Comparisons
+# ============================================================
+
+def multi_group_comparison():
+    """Compare three or more groups."""
+    # Three groups
+    group_a = np.random.normal(10, 2, 40)
+    group_b = np.random.normal(12, 2, 40)
+    group_c = np.random.normal(14, 2, 40)
+
+    # One-way ANOVA
+    result = stx.stats.test_anova(
+        group_a, group_b, group_c,
+        return_as="dataframe"
+    )
+    print("One-way ANOVA:")
+    print(result)
+    # Returns: F, p, eta², omega², power
+
+    # Kruskal-Wallis (non-parametric)
+    result_kw = stx.stats.test_kruskal(
+        group_a, group_b, group_c,
+        return_as="dataframe"
+    )
+    print("\\nKruskal-Wallis:")
+    print(result_kw)
+
+    # Post-hoc tests (after significant ANOVA)
+    result_posthoc = stx.stats.test_tukey_hsd(
+        group_a, group_b, group_c,
+        return_as="dataframe"
+    )
+    print("\\nTukey HSD post-hoc:")
+    print(result_posthoc)
+
+# ============================================================
+# Pattern 3: Correlation Analysis
+# ============================================================
+
+def correlation_analysis():
+    """Analyze relationships between variables."""
+    # Generate correlated data
+    x = np.random.normal(0, 1, 100)
+    y = 0.7 * x + np.random.normal(0, 0.5, 100)
+
+    # Pearson correlation
+    result = stx.stats.test_pearsonr(
+        x, y,
+        return_as="dataframe"
+    )
+    print("Pearson correlation:")
+    print(result)
+    # Returns: r, p, CI, R², power
+
+    # Spearman correlation (non-parametric)
+    result_spearman = stx.stats.test_spearmanr(
+        x, y,
+        return_as="dataframe"
+    )
+    print("\\nSpearman correlation:")
+    print(result_spearman)
+
+# ============================================================
+# Pattern 4: Categorical Data Analysis
+# ============================================================
+
+def categorical_analysis():
+    """Analyze categorical/count data."""
+    # Chi-square test
+    observed = np.array([[30, 20], [15, 35]])
+
+    result = stx.stats.test_chi2(
+        observed,
+        return_as="dataframe"
+    )
+    print("Chi-square test:")
+    print(result)
+    # Returns: chi2, p, Cramer's V, df
+
+    # Fisher's exact test (for small samples)
+    small_table = np.array([[5, 2], [1, 6]])
+    result_fisher = stx.stats.test_fisher_exact(
+        small_table,
+        return_as="dataframe"
+    )
+    print("\\nFisher's exact test:")
+    print(result_fisher)
+
+# ============================================================
+# Pattern 5: With @stx.session (Recommended)
+# ============================================================
+
+@stx.session
+def main(
+    n_subjects=50,
+    CONFIG=stx.INJECTED,
+    plt=stx.INJECTED,
+    COLORS=stx.INJECTED,
+    logger=stx.INJECTED,
+):
+    """Complete statistical analysis workflow."""
+
+    # Generate experimental data
+    control = np.random.normal(100, 15, n_subjects)
+    treatment = np.random.normal(115, 15, n_subjects)
+
+    # 1. Run statistical test
+    result = stx.stats.test_ttest_ind(
+        control, treatment,
+        return_as="dataframe"
+    )
+    logger.info(f"Statistical test completed")
+
+    # 2. Save results
+    stx.io.save(result, CONFIG.SDIR_OUT / "stats_results.csv")
+
+    # 3. Create visualization with stats annotation
+    fig, ax = plt.subplots()
+
+    # Box plot
+    ax.boxplot([control, treatment], labels=["Control", "Treatment"])
+    ax.set_xyt("Group", "Score", "Treatment Effect")
+
+    # Add significance annotation
+    p_value = result["p_value"].values[0]
+    stars = stx.stats.p_to_stars(p_value)
+    ax.annotate(
+        stars,
+        xy=(1.5, max(control.max(), treatment.max()) * 1.05),
+        ha="center", fontsize=14
+    )
+
+    stx.io.save(fig, CONFIG.SDIR_OUT / "treatment_effect.png")
+
+    # 4. Export for paper
+    latex_result = stx.stats.test_ttest_ind(
+        control, treatment,
+        return_as="latex"
+    )
+    stx.io.save(latex_result, CONFIG.SDIR_OUT / "stats_latex.txt")
+
+    logger.info(f"Results saved to {{CONFIG.SDIR_OUT}}")
+    return 0
+
+# ============================================================
+# Pattern 6: Power Analysis & Sample Size
+# ============================================================
+
+def power_analysis_example():
+    """Calculate required sample size or achieved power."""
+    # Calculate required sample size for t-test
+    required_n = stx.stats.power_ttest(
+        effect_size=0.5,    # Medium effect (Cohen's d)
+        alpha=0.05,
+        power=0.80,
+        alternative="two-sided"
+    )
+    print(f"Required sample size per group: {{required_n}}")
+
+    # Calculate achieved power
+    achieved_power = stx.stats.power_ttest(
+        effect_size=0.5,
+        alpha=0.05,
+        n=30,               # Given sample size
+        alternative="two-sided"
+    )
+    print(f"Achieved power with n=30: {{achieved_power:.3f}}")
+
+# ============================================================
+# Available Statistical Tests
+# ============================================================
+
+"""
+stx.stats Available Tests (23 total):
+=====================================
+
+Parametric (Continuous):
+------------------------
+- test_ttest_ind       : Independent samples t-test
+- test_ttest_rel       : Paired samples t-test
+- test_ttest_1samp     : One-sample t-test
+- test_anova           : One-way ANOVA
+- test_anova_rm        : Repeated measures ANOVA
+- test_pearsonr        : Pearson correlation
+
+Non-parametric:
+---------------
+- test_mannwhitneyu    : Mann-Whitney U test
+- test_wilcoxon        : Wilcoxon signed-rank test
+- test_kruskal         : Kruskal-Wallis H test
+- test_friedman        : Friedman test
+- test_spearmanr       : Spearman correlation
+
+Categorical:
+------------
+- test_chi2            : Chi-square test
+- test_fisher_exact    : Fisher's exact test
+- test_mcnemar         : McNemar's test
+
+Post-hoc:
+---------
+- test_tukey_hsd       : Tukey HSD
+- test_dunn            : Dunn's test (after Kruskal)
+
+Normality:
+----------
+- test_shapiro         : Shapiro-Wilk test
+- test_normaltest      : D'Agostino-Pearson test
+
+Helpers:
+--------
+- p_to_stars(p)        : Convert p-value to significance stars
+- power_ttest(...)     : Power analysis for t-tests
+- effect_size_cohens_d : Calculate Cohen's d
+- correct_pvalues      : Multiple comparison correction (FDR, Bonferroni)
+
+CLI Commands:
+-------------
+scitex stats recommend --n-groups 2    # Recommend appropriate test
+scitex stats --help                    # Show all options
+
+MCP Tools:
+----------
+stats_recommend_tests   # Get test recommendations
+stats_run_test         # Execute a specific test
+stats_format_results   # Format for publication
+stats_power_analysis   # Power/sample size calculation
+stats_correct_pvalues  # Multiple comparison correction
+"""
+
+if __name__ == "__main__":
+    main()
+''',
+}
+
+__all__ = ["TEMPLATE"]
+
+# EOF