# Install PyCaret 2.0
!pip install pycaret==2.0

# Import pandas for data loading and manipulation
import pandas as pd

# To render interactive plots in Google Colab
from pycaret.utils import enable_colab
enable_colab()

# Load the data
data = pd.read_csv('penguins.csv')
data.head()

# View the data description
data.info()

from pycaret.classification import *

clf = setup(
    data=data, 
    target='sex',
    train_size=0.8,
    normalize=True,
    session_id=123
)

# Compare and sort models by AUC
compare_models(sort='AUC')

# Get the ID of the models
help(create_model)

# Create an Extra Trees Classifier	
et = create_model('et')

# Tune the classifier	
tuned_et = tune_model(et, optimize='AUC')

tuned_et

plot_model(tuned_et)

plot_model(tuned_et, plot = 'confusion_matrix')

# Precision Recall Curve
plot_model(tuned_et, plot = 'pr')

plot_model(tuned_et, plot = 'class_report')

plot_model(tuned_et, plot='feature')

evaluate_model(tuned_et)

# Interpret the model using SHAP values
interpret_model(tuned_et)

# Make predictions on the test set
predict_model(tuned_et)

# Finalize the model
finalize_model(tuned_et)

# Save the model
save_model(tuned_et, 'et_model_05082020')