# -*- coding: utf-8 -*-
"""Untitled11.ipynb

Automatically generated by Colab.

Original file is located at
    https://colab.research.google.com/drive/1k64KEyZu9c-acfC9LCUOxu3iLV_z0g1o
"""

# Install libraries
!pip install pandas matplotlib seaborn

# Import libraries
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

# Create data directly
data = {
    'Thickness (mm)': [1,1,1,1,1,1,1,1,1,
                       2,2,2,2,2,2,2,2,2,
                       3,3,3,3,3,3,3,3,3],
    'Rotation Speed (Rpm)': [1000,1500,2000,1200,1500,2000,1000,1500,2000,
                             1200,1500,2000,1000,1500,2000,1000,1500,2000,
                             1000,1500,2000,1000,1500,2000,1200,1500,2000],
    'Dwelling Time (s)': [3,3,3,6,6,6,9,9,9,
                          3,3,3,6,6,6,9,9,9,
                          3,3,3,6,6,6,9,9,9],
    'Max Temperature (C)': [74,102,112,93,139,166,125,147,176,
                            85,115,128,96,133,165,132,153,179,
                            97,126,138,118,148,187,136,180,215],
    'Lap-Shear fracture load (KN)': [1.23,1.42,1.61,1.33,1.52,1.73,1.55,2,2.2,
                                     1.2,1.34,1.53,1.65,1.84,2.19,2.26,2.33,2.59,
                                     1.51,1.82,2.11,1.84,2.14,2.35,2.13,2.36,2.54]
}

# Convert data to DataFrame
df = pd.DataFrame(data)

# Display first 5 rows
print("First 5 rows of the data:")
print(df.head())

# Data information
print("\nData Information:")
print(df.info())

# Statistical description
print("\nStatistical Description:")
print(df.describe())

# Plot distribution for each variable
df.hist(bins=20, figsize=(15, 10), edgecolor='black')
plt.suptitle('Distribution of Each Feature', fontsize=20)
plt.show()

# Plot correlation matrix
plt.figure(figsize=(10, 8))
sns.heatmap(df.corr(), annot=True, cmap='coolwarm', fmt=".2f")
plt.title('Correlation Heatmap', fontsize=18)
plt.show()

# Pairplot of variables
sns.pairplot(df, diag_kind='kde')
plt.suptitle('Pairplot of Features', fontsize=20)
plt.show()

# Install libraries
!pip install pandas matplotlib seaborn scikit-learn

# Import libraries
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
from sklearn.linear_model import LinearRegression
from sklearn.metrics import r2_score

# Create data
data = {
    'Thickness (mm)': [1,1,1,1,1,1,1,1,1,
                       2,2,2,2,2,2,2,2,2,
                       3,3,3,3,3,3,3,3,3],
    'Rotation Speed (Rpm)': [1000,1500,2000,1200,1500,2000,1000,1500,2000,
                             1200,1500,2000,1000,1500,2000,1000,1500,2000,
                             1000,1500,2000,1000,1500,2000,1200,1500,2000],
    'Dwelling Time (s)': [3,3,3,6,6,6,9,9,9,
                          3,3,3,6,6,6,9,9,9,
                          3,3,3,6,6,6,9,9,9],
    'Max Temperature (C)': [74,102,112,93,139,166,125,147,176,
                            85,115,128,96,133,165,132,153,179,
                            97,126,138,118,148,187,136,180,215],
    'Lap-Shear fracture load (KN)': [1.23,1.42,1.61,1.33,1.52,1.73,1.55,2,2.2,
                                     1.2,1.34,1.53,1.65,1.84,2.19,2.26,2.33,2.59,
                                     1.51,1.82,2.11,1.84,2.14,2.35,2.13,2.36,2.54]
}

# Convert to DataFrame
df = pd.DataFrame(data)

# Function to compute and display linear regression
def simple_linear_regression(x, y, xlabel, ylabel):
    # reshape data
    X = x.values.reshape(-1,1)
    Y = y.values.reshape(-1,1)

    # create model
    model = LinearRegression()
    model.fit(X, Y)

    # prediction
    Y_pred = model.predict(X)

    # extract coefficients
    slope = model.coef_[0][0]
    intercept = model.intercept_[0]
    r2 = r2_score(Y, Y_pred)

    # print equation
    print(f"\nLinear Regression Equation for {xlabel} vs {ylabel}:")
    print(f"{ylabel} = {slope:.4f} * {xlabel} + {intercept:.4f}")
    print(f"R^2: {r2:.4f}")

    # plot
    plt.figure(figsize=(8,5))
    plt.scatter(X, Y, color='blue', label='Actual Data')
    plt.plot(X, Y_pred, color='red', linewidth=2, label='Regression Line')
    plt.xlabel(xlabel)
    plt.ylabel(ylabel)
    plt.title(f'{ylabel} vs {xlabel}')
    plt.legend()
    plt.grid(True)
    plt.show()

# Compute regressions:
simple_linear_regression(df['Rotation Speed (Rpm)'], df['Lap-Shear fracture load (KN)'],
                         'Rotation Speed (Rpm)', 'Lap-Shear fracture load (KN)')

simple_linear_regression(df['Thickness (mm)'], df['Lap-Shear fracture load (KN)'],
                         'Thickness (mm)', 'Lap-Shear fracture load (KN)')

simple_linear_regression(df['Dwelling Time (s)'], df['Lap-Shear fracture load (KN)'],
                         'Dwelling Time (s)', 'Lap-Shear fracture load (KN)')

# Import additional libraries
from sklearn.preprocessing import StandardScaler

# Standardize data
scaler = StandardScaler()
X_scaled = scaler.fit_transform(X)

# Linear regression on scaled data
model_scaled = LinearRegression()
model_scaled.fit(X_scaled, y)

# Extract coefficients
importance = model_scaled.coef_

# Create importance table
feature_importance = pd.DataFrame({
    'Feature': ['Rotation Speed (Rpm)', 'Dwelling Time (s)', 'Thickness (mm)'],
    'Importance': importance
})

# Sort descending by absolute value
feature_importance = feature_importance.sort_values(by='Importance', key=abs, ascending=False)

print("Feature Importance (sorted):")
print(feature_importance)

# Plot
plt.figure(figsize=(8,5))
sns.barplot(x='Importance', y='Feature', data=feature_importance, palette='viridis')
plt.title('Feature Importance (Linear Regression) - Lap-Shear fracture load')
plt.grid()
plt.show()

# Reload data
data = {
    'Thickness (mm)': [1,1,1,1,1,1,1,1,1,2,2,2,2,2,2,2,2,2,3,3,3,3,3,3,3,3,3],
    'Rotation Speed (Rpm)': [1000,1500,2000,1200,1500,2000,1000,1500,2000,
                              1200,1500,2000,1000,1500,2000,1000,1500,2000,
                              1000,1500,2000,1000,1500,2000,1200,1500,2000],
    'Dwelling Time (s)': [3,3,3,6,6,6,9,9,9,
                          3,3,3,6,6,6,9,9,9,
                          3,3,3,6,6,6,9,9,9],
    'Max Temperature (C)': [74,102,112,93,139,166,125,147,176,
                            85,115,128,96,133,165,132,153,179,
                            97,126,138,118,148,187,136,180,215]
}

df = pd.DataFrame(data)

# inputs and outputs
X = df[['Rotation Speed (Rpm)', 'Dwelling Time (s)', 'Thickness (mm)']]
y = df['Max Temperature (C)']

# Standardize data
scaler = StandardScaler()
X_scaled = scaler.fit_transform(X)

# Linear regression
model_scaled = LinearRegression()
model_scaled.fit(X_scaled, y)

importance = model_scaled.coef_

feature_importance = pd.DataFrame({
    'Feature': ['Rotation Speed (Rpm)', 'Dwelling Time (s)', 'Thickness (mm)'],
    'Importance': importance
})

feature_importance = feature_importance.sort_values(by='Importance', key=abs, ascending=False)

print("Feature Importance (sorted) for Max Temperature:")
print(feature_importance)

plt.figure(figsize=(8,5))
sns.barplot(x='Importance', y='Feature', data=feature_importance, palette='plasma')
plt.title('Feature Importance (Linear Regression) - Max Temperature')
plt.grid()
plt.show()

# Linear regression equation for temperature
model = LinearRegression()
model.fit(X, y)

y_pred = model.predict(X)

intercept = model.intercept_
coefficients = model.coef_

r2 = r2_score(y, y_pred)

print("Multiple Linear Regression Equation for Max Temperature:")
print(f"Max Temperature (C) = {intercept:.4f} + ({coefficients[0]:.4f}) * Rotation Speed (Rpm) + ({coefficients[1]:.4f}) * Dwelling Time (s) + ({coefficients[2]:.4f}) * Thickness (mm)")
print(f"\nR² = {r2:.4f}")

plt.figure(figsize=(8,6))
plt.scatter(y, y_pred, color='blue', edgecolor='k')
plt.plot([y.min(), y.max()], [y.min(), y.max()], '--r', linewidth=2)
plt.xlabel('Actual Max Temperature (C)')
plt.ylabel('Predicted Max Temperature (C)')
plt.title('Actual vs Predicted Max Temperature')
plt.grid(True)
plt.show()

# Machine learning model comparison
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from sklearn.model_selection import train_test_split
from sklearn.metrics import mean_absolute_error, mean_squared_error, r2_score
from sklearn.svm import SVR
from sklearn.ensemble import RandomForestRegressor
from sklearn.linear_model import HuberRegressor, BayesianRidge, ElasticNet
from sklearn.preprocessing import PolynomialFeatures
from sklearn.linear_model import LinearRegression
from sklearn.pipeline import make_pipeline

# Data
data = {
    'Thickness (mm)': [1,1,1,1,1,1,1,1,1,2,2,2,2,2,2,2,2,2,3,3,3,3,3,3,3,3,3],
    'Rotation Speed (Rpm)': [1000,1500,2000,1200,1500,2000,1000,1500,2000,
                              1200,1500,2000,1000,1500,2000,1000,1500,2000,
                              1000,1500,2000,1000,1500,2000,1200,1500,2000],
    'Dwelling Time (s)': [3,3,3,6,6,6,9,9,9,
                          3,3,3,6,6,6,9,9,9,
                          3,3,3,6,6,6,9,9,9],
    'Max Temperature (C)': [74,102,112,93,139,166,125,147,176,
                            85,115,128,96,133,165,132,153,179,
                            97,126,138,118,148,187,136,180,215]
}

df = pd.DataFrame(data)

X = df[['Rotation Speed (Rpm)', 'Dwelling Time (s)', 'Thickness (mm)']]
y = df['Max Temperature (C)']

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)

models = {
    "Support Vector Regression": SVR(),
    "Random Forest Regression": RandomForestRegressor(random_state=42),
    "Robust Regression": HuberRegressor(),
    "Polynomial Regression (Non-Linear)": make_pipeline(PolynomialFeatures(degree=2), LinearRegression()),
    "Bayesian Ridge Regression": BayesianRidge(),
    "Elastic Net Regression": ElasticNet(random_state=42)
}

def evaluate_model(name, model, X_train, X_test, y_train, y_test):
    model.fit(X_train, y_train)
    y_train_pred = model.predict(X_train)
    y_test_pred = model.predict(X_test)

    results = {
        "Model": name,
        "Train MAE": mean_absolute_error(y_train, y_train_pred),
        "Test MAE": mean_absolute_error(y_test, y_test_pred),
        "Train MSE": mean_squared_error(y_train, y_train_pred),
        "Test MSE": mean_squared_error(y_test, y_test_pred),
        "Train RMSE": np.sqrt(mean_squared_error(y_train, y_train_pred)),
        "Test RMSE": np.sqrt(mean_squared_error(y_test, y_test_pred)),
        "Train R2": r2_score(y_train, y_train_pred),
        "Test R2": r2_score(y_test, y_test_pred)
    }
    return results

results = []

for name, model in models.items():
    res = evaluate_model(name, model, X_train, X_test, y_train, y_test)
    results.append(res)

results_df = pd.DataFrame(results)
print(results_df)

results_df.set_index('Model')[['Test R2']].plot(kind='barh', figsize=(10,6), legend=True, title="Test R2 Score by Model")
plt.grid(True)
plt.show()

metrics = ['Test MAE', 'Test MSE', 'Test RMSE', 'Test R2']

for metric in metrics:
    results_df.set_index('Model')[[metric]].plot(kind='barh', figsize=(10,6), legend=True)
    plt.title(f'{metric} Comparison Across Models')
    plt.xlabel(metric)
    plt.ylabel('Model')
    plt.grid(True)
    plt.show()

metrics = ['Test MAE', 'Test MSE', 'Test RMSE', 'Test R2']

for metric in metrics:
    ax = results_df.set_index('Model')[[metric]].plot(kind='barh', figsize=(10,6), legend=True)
    plt.title(f'{metric} Comparison Across Models')
    plt.xlabel(metric)
    plt.ylabel('Model')
    plt.grid(True)

    for i, value in enumerate(results_df[metric]):
        ax.text(value + 0.01, i, f"{value:.3f}", va='center', fontsize=10)

    plt.show()

metrics = ['Test MAE', 'Test MSE', 'Test RMSE', 'Test R2']

for metric in metrics:
    temp_df = results_df.copy()

    if metric == 'Test R2':
        temp_df[metric] = temp_df[metric] * 100
        temp_df['Model_with_value'] = temp_df['Model'] + " (" + temp_df[metric].round(2).astype(str) + "%)"
    else:
        temp_df['Model_with_value'] = temp_df['Model'] + " (" + temp_df[metric].round(3).astype(str) + ")"

    ax = temp_df.set_index('Model_with_value')[[metric]].plot(kind='barh', figsize=(12,7), legend=True)
    plt.title(f'{metric} Comparison Across Models')
    plt.xlabel(metric)
    plt.ylabel('Model')
    plt.grid(True)
    plt.show()