"""
logistic_regression_with_scikit_learn.py
trains and uses a model to predict one of three classes for each input
"""

# Import needed libraries.
import random
from sklearn.datasets import load_iris
from sklearn.linear_model import LogisticRegression

# Set parameters.
number_of_prediction_inputs = 100

# Load test data.
X, y = load_iris(return_X_y=True)
print("X - Data Features:")
print(X)
print("y - Data Classes:")
print(y)

# Instantiate a model.
model = LogisticRegression(random_state=0)

# Train the model.
estimator = model.fit(X, y)

# Get the training score (accuracy).
score = estimator.score(X, y)
print("Score:")
print(score)

# Create shuffled prediction input data.
shuffled_input_data = X
random.shuffle(shuffled_input_data)
print("Shuffled Input Data:")
print(shuffled_input_data)

# Get prediction input data from the shuffled training data.
prediction_input = shuffled_input_data[:number_of_prediction_inputs, :]
print("Prediction Input:")
print(prediction_input)

# Make predictions.
predicted_classes = estimator.predict(prediction_input)
print("Predicted Classes: ")
print(predicted_classes)

# Get prediction probabilities for each class.
probabilities = estimator.predict_proba(prediction_input)
print("Probabilities: ")
print(probabilities)