# -*- coding: utf-8 -*-
"""Email_spam_calsifier

Automatically generated by Colab.

Original file is located at
    https://colab.research.google.com/drive/1obsGCm8CluG_jKB59HlJAlyO5movzKD_

# IMPORT LIBRARIES
"""

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import pickle

from sklearn.model_selection import train_test_split
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.svm import SVC
from sklearn.metrics import accuracy_score, precision_score, recall_score, f1_score
from sklearn.metrics import confusion_matrix, ConfusionMatrixDisplay

"""#**LOAD DATA**"""

df = pd.read_csv('/content/mail_data.csv')

# Replace null values
data = df.where((pd.notnull(df)), '')

# Convert labels
data.loc[data['Category'] == 'spam', 'Category'] = 0
data.loc[data['Category'] == 'ham', 'Category'] = 1

# Separate features and labels
x = data['Message']
y = data['Category'].astype('int')

"""#**TRAIN TEST SPLIT**

"""

x_train, x_test, y_train, y_test = train_test_split(x, y, test_size=0.2, random_state=3)

"""#**TF-IDF**"""

tfidf = TfidfVectorizer(min_df=1, stop_words='english', lowercase=True)

x_train_features = tfidf.fit_transform(x_train)
x_test_features = tfidf.transform(x_test)

"""#**SVM MODEL**"""

model = SVC()

model.fit(x_train_features, y_train)

"""#**PREDICTIONS**"""

train_pred = model.predict(x_train_features)
test_pred = model.predict(x_test_features)

"""#**EVALUATION**"""

train_acc = accuracy_score(y_train, train_pred)
test_acc = accuracy_score(y_test, test_pred)

precision = precision_score(y_test, test_pred)
recall = recall_score(y_test, test_pred)
f1 = f1_score(y_test, test_pred)

print("Training Accuracy:", train_acc)
print("Testing Accuracy:", test_acc)
print("Precision:", precision)
print("Recall:", recall)
print("F1 Score:", f1)

"""#**SAMPLE TEST**"""

sample = ["claim your free gift card today"]

sample_vec = tfidf.transform(sample)
result = model.predict(sample_vec)

print("\nPrediction:", "Ham" if result[0] == 1 else "Spam")

"""#**SAVE MODEL**"""

pickle.dump(model, open("email_model.pkl", "wb"))
pickle.dump(tfidf, open("email_vectorizer.pkl", "wb"))

from google.colab import files
files.download("email_model.pkl")
files.download("email_vectorizer.pkl")

"""#**BAR CHART**"""

plt.figure()
data['Category'].value_counts().plot(kind='bar')
plt.title("Spam vs Ham Emails")
plt.xlabel("Category (0=Spam, 1=Ham)")
plt.ylabel("Count")
plt.savefig("bar_chart.png")
plt.show()

"""#**CONFUSION MATRIX**"""

cm = confusion_matrix(y_test, test_pred)
disp = ConfusionMatrixDisplay(confusion_matrix=cm)
disp.plot()
plt.savefig("confusion_matrix.png")
plt.show()

"""#**ACCURACY COMPARISON**"""

plt.figure()
plt.bar(['Train Accuracy', 'Test Accuracy'], [train_acc, test_acc])
plt.title("Train vs Test Accuracy")
plt.savefig("accuracy_graph.png")
plt.show()

"""#**DOWNLOAD GRAPHS**"""

files.download("bar_chart.png")
files.download("confusion_matrix.png")
files.download("accuracy_graph.png")

!pip install streamlit pyngrok

# Commented out IPython magic to ensure Python compatibility.
# %%writefile app.py
# import streamlit as st
# import pickle
# 
# model = pickle.load(open("email_model.pkl", "rb"))
# tfidf = pickle.load(open("email_vectorizer.pkl", "rb"))
# 
# st.title("Email Spam Detection App")
# 
# user_input = st.text_area("Type your email here:")
# 
# if st.button("Predict"):
#     if user_input.strip() == "":
#         st.warning("Please enter some text!")
#     else:
#         input_features = tfidf.transform([user_input])
#         prediction = model.predict(input_features)
#         st.success("Ham Email " if prediction[0]==1 else "Spam Email ")

!npm install -g localtunnel

!streamlit run app.py & npx localtunnel --port 8501