app.py

import pandas as pd
import plotly.express as px
import streamlit as st
from transformers import pipeline
import datetime
import matplotlib.pyplot as plt

# Function to add background image to the app
def add_bg_from_url(image_url):
    st.markdown(
        f"""
        <style>
        .stApp {{
            background-image: url({image_url});
            background-size: cover;
            background-position: center center;
            background-repeat: no-repeat;
        }}
        </style>
        """,
        unsafe_allow_html=True
    )

# Set background image (it will remain even after file upload)
add_bg_from_url('https://huggingface.co/spaces/engralimalik/Smart-Expense-Tracker/resolve/main/top-view-finance-business-elements.jpg')

# File upload
uploaded_file = st.file_uploader("Upload your expense CSV file", type=["csv"])
if uploaded_file:
    df = pd.read_csv(uploaded_file)

    # Display first few rows to the user for format verification
    st.write("Here are the first few entries in your file for format verification:")
    st.write(df.head())

    # Ensure 'Amount' is numeric and 'Date' is in datetime format
    df['Amount'] = pd.to_numeric(df['Amount'], errors='coerce')
    df['Date'] = pd.to_datetime(df['Date'])

    # Initialize Hugging Face model for zero-shot classification
    classifier = pipeline('zero-shot-classification', model='roberta-large-mnli')
    categories = ["Groceries", "Rent", "Utilities", "Entertainment", "Dining", "Transportation"]

    # Function to categorize
    def categorize_expense(description):
        result = classifier(description, candidate_labels=categories)
        return result['labels'][0]  # Most probable category

    # Apply categorization
    df['Category'] = df['Description'].apply(categorize_expense)

    # Sidebar for setting the monthly budget using sliders
    st.sidebar.header("Set Your Monthly Budget")
    groceries_budget = st.sidebar.slider("Groceries Budget", 0, 1000, 300)
    rent_budget = st.sidebar.slider("Rent Budget", 0, 5000, 1000)
    utilities_budget = st.sidebar.slider("Utilities Budget", 0, 500, 150)
    entertainment_budget = st.sidebar.slider("Entertainment Budget", 0, 1000, 100)
    dining_budget = st.sidebar.slider("Dining Budget", 0, 1000, 150)
    transportation_budget = st.sidebar.slider("Transportation Budget", 0, 500, 120)

    # Store the updated budget values
    budgets = {
        "Groceries": groceries_budget,
        "Rent": rent_budget,
        "Utilities": utilities_budget,
        "Entertainment": entertainment_budget,
        "Dining": dining_budget,
        "Transportation": transportation_budget
    }

    # Get the minimum and maximum dates from the uploaded file
    min_date = df['Date'].min()
    max_date = df['Date'].max()

    # Add a date slider for start and end date based on data from the uploaded file
    start_date = st.sidebar.date_input("Start Date", min_date)
    end_date = st.sidebar.date_input("End Date", max_date)

    # Filter data by date range
    df_filtered = df[(df['Date'] >= pd.to_datetime(start_date)) & (df['Date'] <= pd.to_datetime(end_date))]

    # Track if any category exceeds its budget
    df_filtered['Budget_Exceeded'] = df_filtered.apply(lambda row: row['Amount'] > budgets.get(row['Category'], 0), axis=1)

    # Show categories that exceeded their budget
    exceeded_budget = df_filtered[df_filtered['Budget_Exceeded'] == True]
    st.write("Categories that exceeded the budget:", exceeded_budget[['Date', 'Category', 'Amount']])

    # Visualizations

    # 1. Pie Chart for expense distribution by category
    category_expenses = df_filtered.groupby('Category')['Amount'].sum()
    fig1 = px.pie(category_expenses, values=category_expenses.values, names=category_expenses.index, title="Expense Distribution by Category")
    st.plotly_chart(fig1)

    # 2. Monthly Spending Trends (Line Chart)
    df_filtered['Month'] = df_filtered['Date'].dt.to_period('M').astype(str)  # Convert Period to string for Plotly compatibility
    monthly_expenses = df_filtered.groupby('Month')['Amount'].sum()

    # Convert monthly_expenses into DataFrame for correct plotting
    monthly_expenses_df = monthly_expenses.reset_index()
    if not monthly_expenses_df.empty:
        fig2 = px.line(monthly_expenses_df, x='Month', y='Amount', title="Monthly Expenses", labels={"Month": "Month", "Amount": "Amount ($)"})
        st.plotly_chart(fig2)
    else:
        st.write("No data to display for the selected date range.")

    # 3. Monthly Spending vs Budget (Bar Chart)
    if not monthly_expenses_df.empty:
        monthly_expenses_df = pd.DataFrame({
            'Actual': monthly_expenses,
            'Budget': [sum(budgets.values())] * len(monthly_expenses)  # Same budget for simplicity
        })
        
        # Create a Matplotlib figure for the bar chart
        fig3, ax = plt.subplots(figsize=(10, 6))
        monthly_expenses_df.plot(kind='bar', ax=ax)
        ax.set_title('Monthly Spending vs Budget')
        ax.set_ylabel('Amount ($)')
        ax.set_xlabel('Month')

        # Use st.pyplot to display the figure
        st.pyplot(fig3)
    else:
        st.write("No data to display for the selected date range.")