ML - Lab 3 - Categorical NB, Laplacian Smoothing

import pandas as pd
from matplotlib.pyplot import plot as plt
from sklearn.preprocessing import OneHotEncoder
from sklearn.naive_bayes import CategoricalNB


# NAIVE BAYES
# Load data
traffic = pd.read_csv("./traffic.txt")
X = traffic.loc[:, ["Weather", "Day"]]
y = traffic.loc[:, "HighTraffic"]
print()
print("****************************************************")
print("**** NAIVE BAYES ****")
print("Traffic = ")
print(traffic)
print()

# Encoder
encoder = OneHotEncoder(handle_unknown="ignore", sparse=False)
encoder = encoder.fit(X)
X = encoder.transform(X)

# Classification with Naive-Bayes
clf = CategoricalNB(alpha=0.0001)
clf.fit(X, y)

# Prediction
new1 = {"Weather": ["Hot"], "Day": ["Vacation"]}
new_data = pd.DataFrame(new1)
transformed_new_data = encoder.transform(new_data)
print("New1 = " + str(new1))
print("Prediction = " + str(clf.predict(transformed_new_data)))
print("Probabilities = " + str(clf.predict_proba(transformed_new_data)))
print()
new2 = {"Weather": ["Hot"], "Day": ["Weekend"]}
new_data2 = pd.DataFrame(new2)
transformed_new_data2 = encoder.transform(new_data2)
print("New2 = " + str(new2))
print("Prediction = " + str(clf.predict(transformed_new_data2)))
print("Probabilities = " + str(clf.predict_proba(transformed_new_data2)))
print()
print()
print()
print()




# LAPLACE SMOOTHING
print("****************************************************")
print("**** LAPLACE SMOOTHING ****")

# Classification with Naive-Bayes
clf = CategoricalNB(alpha=0.9999)
clf.fit(X, y)
new2 = {"Weather": ["Hot"], "Day": ["Weekend"]}
new_data2 = pd.DataFrame(new2)
transformed_new_data2 = encoder.transform(new_data2)
print("New2 = " + str(new2))
print("Prediction = " + str(clf.predict(transformed_new_data2)))
print("Probabilities = " + str(clf.predict_proba(transformed_new_data2)))