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

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

churn_counts = df_satisfaction['Churn'].value_counts()
print('Churned vs Not Churned:')
print(churn_counts)

contract_counts = df_satisfaction['Contract'].value_counts()
print('Customers by Contract Type:')
print(contract_counts)

df_satisfaction['Segment'] = np.where(df_satisfaction['tenure'] < 12, 'New', 'Established')
print(df_satisfaction[['tenure','Segment']].head(5))

churn_by_segment = pd.crosstab(df_satisfaction['Segment'], df_satisfaction['Churn'], normalize='index')
print(churn_by_segment)

url = 'https://archive.ics.uci.edu/ml/machine-learning-databases/00502/online_retail_II.xlsx'
df_retail = pd.read_excel(url, sheet_name='Year 2010-2011')
df_retail['InvoiceDate'] = pd.to_datetime(df_retail['InvoiceDate'])
print(df_retail.shape)
print(df_retail[['Customer ID','InvoiceDate','Country']].head(3))

customers_by_country = df_retail.groupby('Country')['Customer ID'].nunique().sort_values(ascending=False)
print(customers_by_country.head(5))

customer_value = df_retail.groupby('Customer ID')['Quantity'].sum()
value_threshold = customer_value.quantile(0.9)
high_value_ids = customer_value[customer_value > value_threshold].index
df_retail['ValueSegment'] = np.where(df_retail['Customer ID'].isin(high_value_ids), 'High Value', 'Regular')
print(df_retail[['Customer ID','Quantity','ValueSegment']].drop_duplicates().head(7))

df_satisfaction['Segment2'] = np.select(
    [
        (df_satisfaction['tenure'] < 12) & (df_satisfaction['Contract']=='Month-to-month'),
        (df_satisfaction['tenure'] >= 12) & (df_satisfaction['Contract']=='Two year')
    ],
    ['At-Risk New', 'Loyal Long-Term'],
    default='Other'
)
print(df_satisfaction[['tenure','Contract','Segment2']].head(7))

multi_seg_churn = pd.crosstab(df_satisfaction['Segment2'], df_satisfaction['Churn'], normalize='index')
print(multi_seg_churn)

np.random.seed(42)
df_nps = pd.DataFrame({
    'CustomerID': range(1,501),
    'Age': np.random.randint(18,70,500),
    'Region': np.random.choice(['North','South','East','West'],500),
    'NPS_Score': np.random.randint(0,11,500)
})
print(df_nps.shape)
print(df_nps.head(3))

def nps_segment(score):
    if score <= 6:
        return 'Detractor'
    elif score <= 8:
        return 'Passive'
    else:
        return 'Promoter'
df_nps['NPS_Segment'] = df_nps['NPS_Score'].apply(nps_segment)
print(df_nps.groupby('NPS_Segment').size())

nps_by_region = pd.crosstab(df_nps['Region'], df_nps['NPS_Segment'], normalize='index')
print(nps_by_region.round(2))

summary = df_nps.groupby(['Region','NPS_Segment']).size().unstack().fillna(0)
summary.to_csv('nps_segment_summary.csv')
print('Segmentation summary file saved as nps_segment_summary.csv')

missing = df_retail.isnull().sum()
print('Missing values per column:')
print(missing[missing > 0])

try:
    result = df_retail.groupby('ValuSegment')['Customer ID'].nunique()
except KeyError as e:
    print('Error:', e)
    print('Check for correct segment column name: should be ValueSegment')

def safe_nps_segment(score):
    if pd.isnull(score) or not (0 <= score <= 10):
        return 'Unknown'
    elif score <= 6:
        return 'Detractor'
    elif score <= 8:
        return 'Passive'
    else:
        return 'Promoter'
df_nps['Safe_NPS_Segment'] = df_nps['NPS_Score'].apply(safe_nps_segment)
print(df_nps['Safe_NPS_Segment'].value_counts())

tenure_contract_ct = pd.crosstab(df_satisfaction['Segment'], df_satisfaction['Contract'])
print(tenure_contract_ct)

df_satisfaction['LoyaltyIndex'] = df_satisfaction['tenure'] * (df_satisfaction['Contract']=='Two year').astype(int)
print(df_satisfaction[['tenure','Contract','LoyaltyIndex']].head(6))

df_retail['Month'] = df_retail['InvoiceDate'].dt.to_period('M')
monthly_counts = df_retail.groupby('Month')['Customer ID'].nunique()
print(monthly_counts.tail(12))

segment_counts = df_satisfaction['Segment2'].value_counts()
at_risk_churn = multi_seg_churn.loc['At-Risk New', 'Yes'] if 'At-Risk New' in multi_seg_churn.index else None
if at_risk_churn is not None and at_risk_churn > 0.3:
    recommendation = 'Launch onboarding and specials for At-Risk New to reduce churn.'
else:
    recommendation = 'Focus on other segments or maintain current actions.'
print('Segment sizes:')
print(segment_counts)
print('At-Risk New churn rate:', at_risk_churn)
print('Recommendation:', recommendation)