Test

import numpy as np
import pandas as pd
from sklearn.ensemble import IsolationForest
from sklearn.preprocessing import StandardScaler, OneHotEncoder
from sklearn.compose import ColumnTransformer

# Assuming `df` is your original DataFrame

# Step 1: Separate categorical and numerical columns
categorical_cols = df.select_dtypes(include=['object', 'category']).columns
numerical_cols = df.select_dtypes(include=['float32', 'float64', 'int32', 'int64']).columns

# Step 2: Preprocessing pipeline
# One-hot encode categorical columns and scale numerical columns
preprocessor = ColumnTransformer(
    transformers=[
        ('num', StandardScaler(), numerical_cols),
        ('cat', OneHotEncoder(drop='first'), categorical_cols)
    ])

# Apply preprocessing (scaling and encoding)
X_processed = preprocessor.fit_transform(df)

# Step 3: Convert preprocessed data to a DataFrame (optional, for easier handling)
# Get feature names from the preprocessor for encoded columns
encoded_cols = preprocessor.named_transformers_['cat'].get_feature_names_out(categorical_cols)
all_columns = np.concatenate([numerical_cols, encoded_cols])

# Convert to DataFrame for better handling
X_df = pd.DataFrame(X_processed, columns=all_columns)

# Step 4: Apply Isolation Forest on the processed data
isolation_forest = IsolationForest(n_estimators=100, contamination=0.01, random_state=42)
labels = isolation_forest.fit_predict(X_df)

# Step 5: Merge the results back with the original DataFrame
df_with_labels = df.copy()
df_with_labels['labels'] = labels

# Create a new column that marks anomalies
df_with_labels['is_anomaly'] = df_with_labels['labels'].apply(lambda x: 'Anomaly' if x == -1 else 'Normal')

# Now `df_with_labels` contains the original data, the anomaly labels, and the flag for anomalies



import matplotlib.pyplot as plt
from sklearn.decomposition import PCA

# Reduce data to 2D for visualization using PCA
pca = PCA(n_components=2)
reduced_data = pca.fit_transform(X_processed)

# Create a scatter plot with anomalies in red and normal points in blue
plt.figure(figsize=(10, 6))
plt.scatter(reduced_data[:, 0], reduced_data[:, 1], c=df_with_labels['is_anomaly'].map({'Anomaly': 'red', 'Normal': 'blue'}))
plt.title('Isolation Forest: Anomalies in Red, Normal Points in Blue')
plt.xlabel('PCA Component 1')
plt.ylabel('PCA Component 2')
plt.show()