qdesc 0.1.7__py3-none-any.whl → 0.1.8__py3-none-any.whl

qdesc/__init__.py

@@ -72,6 +72,7 @@ def freqdist_a(df, ascending=False):
     return final_df
 
 def clean_sheet_name(name):
+    import re
     # Remove invalid characters
     name = re.sub(r'[:\\/?*\[\]]', '', name)
     # Limit to 31 characters
@@ -105,3 +106,51 @@ def freqdist_to_excel(df, output_path, sort_by='Percentage', ascending=False, to
             used_names.add(sheet_name.lower())
             distribution.to_excel(writer, sheet_name=sheet_name, index=False)
     print(f"Frequency distributions written to {output_path}")
+
+def normcheck_dashboard(df, significance_level=0.05, figsize=(18, 5)):
+    import pandas as pd
+    import numpy as np
+    import matplotlib.pyplot as plt
+    import seaborn as sns
+    import statsmodels.api as sm
+    from scipy.stats import anderson
+    import math
+    numeric_cols = df.select_dtypes(include=[np.number]).columns
+    if len(numeric_cols) == 0:
+        print("No numeric columns to analyze.")
+        return
+    for col in numeric_cols:
+        data = df[col].dropna()
+        print(f"\n--- Variable: {col} ---")
+        if len(data) < 8:
+            print("Not enough data to perform Anderson-Darling test or meaningful plots.")
+            continue
+        # Anderson-Darling Test
+        test_result = anderson(data, dist='norm')
+        stat = test_result.statistic
+        sig_levels = test_result.significance_level
+        crit_values = test_result.critical_values
+        level_diff = [abs(sl - (significance_level * 100)) for sl in sig_levels]
+        closest_index = level_diff.index(min(level_diff))
+        used_sig = sig_levels[closest_index]
+        crit_val = crit_values[closest_index]
+        decision = "Fail to Reject Null" if stat <= crit_val else "Reject Null"
+        # Print Summary
+        print(f"  Anderson-Darling Statistic : {stat:.4f}")
+        print(f"  Critical Value (@ {used_sig}%) : {crit_val:.4f}")
+        print(f"  Decision : {decision}")
+        # Plots (QQ, Histogram, Boxplot)
+        fig, axes = plt.subplots(1, 3, figsize=figsize)
+        # QQ Plot
+        sm.qqplot(data, line='s', ax=axes[0])
+        axes[0].set_title(f"QQ Plot - {col}")
+        # Histogram (No KDE)
+        sns.histplot(data, bins=30, kde=False, color='gray', alpha=0.3, ax=axes[1])
+        axes[1].set_title(f"Histogram - {col}")
+        # Boxplot
+        sns.boxplot(x=data, ax=axes[2], color='lightblue')
+        axes[2].set_title(f"Boxplot - {col}")
+        axes[2].set_xlabel(col)
+        plt.suptitle(f"Normality Assessment - {col}", fontsize=14, y=1.05)
+        plt.tight_layout()
+        plt.show()

{qdesc-0.1.7.dist-info → qdesc-0.1.8.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: qdesc
-Version: 0.1.7
+Version: 0.1.8
 Summary: Quick and Easy way to do descriptive analysis.
 Author: Paolo Hilado
 Author-email: datasciencepgh@proton.me

qdesc-0.1.8.dist-info/RECORD

@@ -0,0 +1,6 @@
+qdesc/__init__.py,sha256=YtYahB-neaYOG3DvVXweSuFExqVZFNR0lAivaPp9_SA,6599
+qdesc-0.1.8.dist-info/LICENCE.txt,sha256=xdFo-Rt6I7EP7C_qrVeIBIcH_7mRGUh8sciJs2R8VmY,9684
+qdesc-0.1.8.dist-info/METADATA,sha256=k7IeXnqHYiaoz999aax8qWGlOD2khStdVVSRrRzFmsc,3780
+qdesc-0.1.8.dist-info/WHEEL,sha256=2wepM1nk4DS4eFpYrW1TTqPcoGNfHhhO_i5m4cOimbo,92
+qdesc-0.1.8.dist-info/top_level.txt,sha256=JuSs1wWRGN77DVuq-SX-5P7m_mIZF0ikEVgPTBOrHb0,6
+qdesc-0.1.8.dist-info/RECORD,,

qdesc-0.1.7.dist-info/RECORD

@@ -1,6 +0,0 @@
-qdesc/__init__.py,sha256=zmaI_ancCfg7Vuy8opqg8H9j6iwCElfIbGwBM15OJPU,4498
-qdesc-0.1.7.dist-info/LICENCE.txt,sha256=xdFo-Rt6I7EP7C_qrVeIBIcH_7mRGUh8sciJs2R8VmY,9684
-qdesc-0.1.7.dist-info/METADATA,sha256=RtfEWvWuIvU9BxXASizXho7YTYjMt_jzDEWCzfuXZZI,3780
-qdesc-0.1.7.dist-info/WHEEL,sha256=2wepM1nk4DS4eFpYrW1TTqPcoGNfHhhO_i5m4cOimbo,92
-qdesc-0.1.7.dist-info/top_level.txt,sha256=JuSs1wWRGN77DVuq-SX-5P7m_mIZF0ikEVgPTBOrHb0,6
-qdesc-0.1.7.dist-info/RECORD,,