import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
from sklearn.preprocessing import LabelEncoder
from sklearn.model_selection import train_test_split
from sklearn.pipeline import Pipeline
from sklearn.compose import ColumnTransformer
from sklearn.linear_model import LogisticRegression
from sklearn.ensemble import RandomForestClassifier
from sklearn.ensemble import GradientBoostingClassifier
from sklearn.metrics import confusion_matrix, classification_report
import pickle
from pathlib import Path
from jcopml.pipeline import num_pipe, cat_pipe
from jcopml.utils import save_model, load_model
from jcopml.plot import plot_missing_value
from jcopml.feature_importance import mean_score_decrease
df = pd.read_csv("https://dqlab-dataset.s3-ap-southeast-1.amazonaws.com/dqlab_telco_final.csv")
#Tampilkan bentuk dari dataset
print(df.shape)
#Tampilkan 5 data teratas
print(df.head())
#Tampilkan jumlah ID yang unik
print(df.customerID.nunique())
fig = plt.figure()
ax = fig.add_axes([0,0,1,1])
ax.axis('equal')
labels = ['No','Yes']
churn = df.Churn.value_counts()
ax.pie(churn, labels=labels, autopct='%.0f%%')
plt.title("Presentase Churn", fontsize=20)
plt.show()
numerical_features = ['tenure', 'MonthlyCharges', 'TotalCharges']
fig, ax = plt.subplots(1, 3, figsize=(15, 6))
# Use the following code to plot two overlays of histogram per each numerical_features, use a color of blue and orange, respectively
df[df.Churn == 'No'][numerical_features].hist(bins=20, color='blue', alpha=0.5, ax=ax)
df[df.Churn == 'Yes'][numerical_features].hist(bins=20, color='orange', alpha=0.5, ax=ax)
plt.show()
sns.set(style='darkgrid')
# Your code goes here
fig, ax = plt.subplots(3, 3, figsize=(14, 12))
sns.countplot(data=df, x='gender', hue='Churn', ax=ax[0][0])
sns.countplot(data=df, x='Partner', hue='Churn', ax=ax[0][1])
sns.countplot(data=df, x='SeniorCitizen', hue='Churn', ax=ax[0][2])
sns.countplot(data=df, x='PhoneService', hue='Churn', ax=ax[1][0])
sns.countplot(data=df, x='StreamingTV', hue='Churn', ax=ax[1][1])
sns.countplot(data=df, x='InternetService', hue='Churn', ax=ax[1][2])
sns.countplot(data=df, x='PaperlessBilling', hue='Churn', ax=ax[2][1])
plt.tight_layout()
plt.show()
cleaned_df=df.drop(columns=["customerID","UpdatedAt"], axis=1)
print(cleaned_df)
from sklearn.preprocessing import LabelEncoder
#Convert all the non-numeric columns to numerical data types
for column in cleaned_df.columns:
if cleaned_df[column].dtype == np.number: continue
#Perform encoding for each non-numeric column
cleaned_df[column] = LabelEncoder().fit_transform(cleaned_df[column])
print(cleaned_df.describe())
