import pandas as pd
import numpy as np
import openml
import warnings
warnings.filterwarnings('ignore')

dataset = openml.datasets.get_dataset(42178)
df, _, _, _ = dataset.get_data(dataset_format='dataframe')
print(df.shape)
print(df.head(3))

gender_counts = df['gender'].value_counts()
print(gender_counts)

senior_pct = df['SeniorCitizen'].value_counts(normalize=True) * 100
print(senior_pct.round(2))

partner_counts = df['Partner'].value_counts()
print(partner_counts)
dependents_counts = df['Dependents'].value_counts()
print(dependents_counts)

age_groups = pd.cut(df['tenure'], bins=[0, 12, 36, 72], labels=['<1yr', '1-3yrs', '3-6yrs'])
df['TenureGroup'] = age_groups
grouped_tenure = df.groupby('TenureGroup').size()
print(grouped_tenure)

churn_by_gender = df.groupby('gender')['Churn'].value_counts().unstack().fillna(0)
print(churn_by_gender)

avg_monthly_by_senior = df.groupby('SeniorCitizen')['MonthlyCharges'].mean()
print(avg_monthly_by_senior.round(2))

contract_churn = df.groupby('Contract')['Churn'].value_counts(normalize=True).unstack().fillna(0) * 100
print(contract_churn.round(1))

internet_by_senior = pd.crosstab(df['SeniorCitizen'], df['InternetService'])
print(internet_by_senior)

churn_family = df.groupby(['Partner', 'Dependents'])['Churn'].value_counts(normalize=True).unstack().fillna(0) * 100
print(churn_family.round(2))

dataset2 = openml.datasets.get_dataset(1461)
df2, _, _, _ = dataset2.get_data(dataset_format='dataframe')
df2.columns = ['age','job','marital','education','default','balance','housing','loan','contact','day','month','duration','campaign','pdays','previous','poutcome','response']
age_brackets = pd.cut(df2['age'], bins=[15,25,35,50,100], labels=['16-25','26-35','36-50','51+'])
segment_counts = age_brackets.value_counts()
print(segment_counts)

crosstab = pd.crosstab([df2['education'], age_brackets], df2['response'])
print(crosstab)

df2['balance_cat'] = pd.cut(df2['balance'], bins=[-np.inf,0,1000,5000,np.inf], labels=['Debt','Low','Medium','High'])
response_by_segment = df2.groupby(['job','balance_cat'])['response'].value_counts(normalize=True).unstack().fillna(0) * 100
print(response_by_segment.head(6).round(1))

missing_ages = df['tenure'].isnull().sum()
print(f'Missing tenure entries: {missing_ages}')

if 'region' in df.columns:
    print(df['region'].value_counts())
else:
    print('No region field available in this dataset.')

# Example: Incorrect groupings due to a typo
try:
    wrong_seg = df.groupby('Gennder').size()
except Exception as e:
    print(f'Error: {e}')

n_na = df['MonthlyCharges'].isnull().sum()
if n_na > 0:
    print(f'Found {n_na} missing values in MonthlyCharges. Consider filling with median.')
    df['MonthlyCharges'] = df['MonthlyCharges'].fillna(df['MonthlyCharges'].median())
else:
    print('No missing values detected.')

# Example: Misinterpretation of NPS-like score
if 'Churn' in df.columns:
    nps_proxy = (df['Churn']=='No').mean() * 100
    print(f'Percent customers satisfied / loyal (proxy): {nps_proxy:.2f}%')
else:
    print('NPS field unavailable; use a satisfied/loyalty proxy.')

cross_demo = pd.crosstab(df['gender'], df['SeniorCitizen'], margins=True)
print(cross_demo)

df['FamilySegment'] = (df['Partner']=='Yes').astype(str) + '_' + (df['Dependents']=='Yes').astype(str)
result = df.groupby('FamilySegment')['MonthlyCharges'].mean()
print(result)

churn_index = df.groupby('Contract')['Churn'].apply(lambda x: (x=='Yes').mean()*100)
print(churn_index.round(2))

trends = df.groupby('SeniorCitizen')['tenure'].mean()
print(trends.round(1))

np.random.seed(42)
edemo = pd.DataFrame({
    'CustomerID': range(1,201),
    'Gender': np.random.choice(['Male','Female'],200),
    'AgeGroup': np.random.choice(['18-25','26-35','36-50','51+'],200,p=[0.2,0.3,0.3,0.2]),
    'NPS': np.random.randint(0,11,200)
})
print(edemo.head(3))

nps_segment = edemo.groupby(['Gender','AgeGroup'])['NPS'].mean().unstack()
print(nps_segment.round(2))

# Business insight: flag weak segments for attention
min_nps = nps_segment.min().min()
if min_nps < 5:
    print('ALERT: A demographic segment has a low NPS!')
else:
    print('All segments are reasonably satisfied.')