1import pandas as pd
2import numpy as np
3import seaborn as sns
4import matplotlib.pyplot as plt
5from sklearn.model_selection import cross_val_score
6from sklearn.linear_model import LogisticRegression
7from sklearn.naive_bayes import GaussianNB
8from sklearn.neighbors import KNeighborsClassifier
9
10def data_summary(df):
11    # Summary statistics
12    summary_stats = df.describe()
13
14    # Count of missing values
15    missing_values = df.isnull().sum()
16
17    # Count of outliers
18    outliers = {}
19    for col in df.columns:
20        if df[col].dtype != 'object':
21            q1 = df[col].quantile(0.25)
22            q3 = df[col].quantile(0.75)
23            iqr = q3 - q1
24            lower_bound = q1 - 1.5 * iqr
25            upper_bound = q3 + 1.5 * iqr
26            num_outliers = len(df[(df[col] < lower_bound) | (df[col] > upper_bound)])
27            outliers[col] = num_outliers
28
29    # Other measures
30    measures = {}
31    for col in df.columns:
32        unique_values = df[col].nunique()
33        zero_values = len(df[df[col] == 0])
34        measures[col] = {'Unique Values': unique_values, 'Zero Values': zero_values}
35            
36    combined_table = pd.DataFrame(pd.concat([summary_stats.T, pd.Series(missing_values, name='Missing'), pd.Series(outliers, name='Outliers')], axis=1))
37
38    return combined_table, measures