app.py

import streamlit as st
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
import folium
from folium.plugins import HeatMap
from streamlit_folium import st_folium
import plotly.express as px
from datetime import datetime

# Set page config
st.set_page_config(page_title="Nuisance Complaints Dashboard", layout="wide")

# Title and introduction
st.title("Nuisance Complaints Analysis Dashboard")
st.markdown("""
**Team Members:** 
* Lu Chang (luchang2@illinois.edu)
* Qiming Li (qimingl4@illinois.edu)
* Ruchita Alate (ralate2@illinois.edu)
* Shreyas Kulkarni (ssk16@illinois.edu)
* Vishal Devulapalli (nsd3@illinois.edu)
""")
st.write("This dashboard analyzes nuisance complaints data from the City of Urbana.")

# Load and clean data
@st.cache_data
def load_and_clean_data():
    try:
        # Load data
        data = pd.read_csv('Nuisance_Complaints.csv')
        
        # Drop rows with missing 'File Number'
        data = data.dropna(subset=['File Number'])
        
        # Convert dates and handle date-related columns
        data['Date Reported'] = pd.to_datetime(data['Date Reported'])
        data['Date Notice Mailed or Given'] = pd.to_datetime(data['Date Notice Mailed or Given'])
        data['File Close Date'] = pd.to_datetime(data['File Close Date'], errors='coerce')
        
        # Handle 'Date Notice Mailed or Given'
        median_delay = (data['Date Notice Mailed or Given'] - data['Date Reported']).dt.days.median()
        data.loc[data['Date Notice Mailed or Given'].isna(), 'Date Notice Mailed or Given'] = \
            data.loc[data['Date Notice Mailed or Given'].isna(), 'Date Reported'] + pd.Timedelta(days=median_delay)
        
        # Handle 'Type of Complaint'
        data['Type of Complaint'] = data['Type of Complaint'].fillna('Unknown')
        
        # Handle 'Disposition'
        most_common_disposition = data.groupby('Type of Complaint')['Disposition'].agg(
            lambda x: x.mode().iloc[0] if len(x.mode()) > 0 else 'Pending'
        )
        data['Disposition'] = data.apply(
            lambda row: most_common_disposition[row['Type of Complaint']] 
            if pd.isna(row['Disposition']) else row['Disposition'], 
            axis=1
        )
        
        # Calculate processing time for resolved cases
        data['Processing Time'] = np.where(
            data['File Close Date'].notna(),
            (data['File Close Date'] - data['Date Reported']).dt.days,
            np.nan
        )
        
        # Handle 'Method Submitted'
        data.loc[
            (data['Submitted Online?']) & (data['Method Submitted'].isna()),
            'Method Submitted'
        ] = 'Online'
        data['Method Submitted'] = data['Method Submitted'].fillna(data['Method Submitted'].mode()[0])
        
        # Drop rows with missing critical values
        data = data.dropna(subset=['Submitted Online?', 'Mapped Location'])
        
        # Extract and clean location data
        data['Latitude'] = data['Mapped Location'].str.extract(r'\(([^,]+),')[0].astype(float)
        data['Longitude'] = data['Mapped Location'].str.extract(r', ([^,]+)\)')[0].astype(float)
        
        # Ensure Year Reported is integer
        data['Year Reported'] = data['Year Reported'].astype(int)
        
        return data
        
    except Exception as e:
        st.error(f"Error in data preprocessing: {str(e)}")
        raise e

# Load the data
try:
    data = load_and_clean_data()
    st.success("Data successfully loaded and cleaned!")
except Exception as e:
    st.error(f"Error loading data: {str(e)}")
    st.stop()

# Create sidebar

st.sidebar.header("Dashboard Controls")

# Get unique years and convert to list for selectbox
year_list = sorted(data['Year Reported'].unique().tolist())
year_options = ['All Time'] + [int(year) for year in year_list]  # Convert years to integers

selected_year = st.sidebar.selectbox(
    "Select Year",
    options=year_options,
)
# Add visualization type selector
viz_type = st.sidebar.selectbox(
    "Select Visualization",
    ["Complaint Types", "Geographic Distribution", "Resolution Status",
     "Submission Methods", "Complaints by Disposition"]
)


# Filter data based on selected year
if selected_year == 'All Time':
    filtered_data = data  # Use complete dataset when 'All Time' is selected
else:
    filtered_data = data[data['Year Reported'] == selected_year]

# Update header text
if selected_year == 'All Time':
    st.header("Analysis for All Time")
else:
    st.header(f"Analysis for Year {selected_year}")
# Main content

# Create metrics
# Create metrics
# Create metrics
# Create metrics
# Create metrics
col1, col2, col3 = st.columns(3)
with col1:
    st.metric("Total Complaints", len(filtered_data))
with col2:
    avg_time = filtered_data['Processing Time'].mean()
    st.metric("Average Processing Time", f"{avg_time:.1f} days" if pd.notna(avg_time) else "N/A")
with col3:
    if not filtered_data.empty:
        most_common = filtered_data['Type of Complaint'].value_counts().index[0]
        st.metric("Most Common Type", most_common)
    else:
        st.metric("Most Common Type", "N/A")
if viz_type == "Complaint Types":
    # Interactive Pie Chart
    st.subheader("Interactive Complaint Types Pie Chart")
    complaint_counts = filtered_data['Type of Complaint'].value_counts().reset_index()
    complaint_counts.columns = ['Complaint Type', 'Count']

    fig = px.pie(
        complaint_counts,
        names='Complaint Type',
        values='Count',
        title=f'Complaint Types Distribution in {selected_year}',
        hole=0.4  # Donut style
    )
    fig.update_traces(textinfo='percent+label')
    st.plotly_chart(fig, use_container_width=True)

elif viz_type == "Geographic Distribution":
    # Clustered Heatmap
    st.subheader("Clustered Heatmap of Complaints")
    map_center = [filtered_data['Latitude'].mean(), filtered_data['Longitude'].mean()]
    m = folium.Map(location=map_center, zoom_start=12)

    heat_data = filtered_data[['Latitude', 'Longitude']].dropna().values.tolist()
    HeatMap(heat_data).add_to(m)

    st_data = st_folium(m, width=700, height=500)


elif viz_type == "Resolution Status":
    st.subheader("Complaint Resolution Status")
    fig, ax = plt.subplots(figsize=(10, 6))
    resolution_counts = filtered_data['Disposition'].value_counts()
    sns.barplot(x=resolution_counts.values, y=resolution_counts.index)
    plt.title(f'Resolution Status Distribution in {selected_year}')
    st.pyplot(fig)

elif viz_type == "Submission Methods":
    st.subheader("Submission Methods Analysis")
    fig, ax = plt.subplots(figsize=(10, 6))
    submission_counts = filtered_data['Method Submitted'].value_counts()
    sns.barplot(x=submission_counts.values, y=submission_counts.index)
    plt.title(f'Submission Methods in {selected_year}')
    st.pyplot(fig)


elif viz_type == "Complaints by Disposition":
    st.subheader("Complaints by Disposition")
    disposition_counts = filtered_data['Disposition'].value_counts()
    
    if not disposition_counts.empty:
        fig, ax = plt.subplots(figsize=(10, 6))
        sns.barplot(x=disposition_counts.values, y=disposition_counts.index, palette="viridis", ax=ax)
        ax.set_title(f'Complaints by Disposition in {selected_year}', fontsize=14)
        ax.set_xlabel('Number of Complaints', fontsize=12)
        ax.set_ylabel('Disposition', fontsize=12)
        st.pyplot(fig)
    else:
        st.write("No data available for the selected year.")

# Additional insights
st.header("Key Insights")
col1, col2 = st.columns(2)

with col1:
    st.subheader("Top 3 Complaint Types")
    top_complaints = filtered_data['Type of Complaint'].value_counts().head(3)
    st.write(top_complaints)

with col2:
    st.subheader("Resolution Efficiency")
    resolution_rate = (filtered_data['Disposition'].value_counts() /
                      len(filtered_data) * 100).round(2)
    st.write(resolution_rate)

# Footer
st.markdown("---")
st.markdown("Dataset provided by the City of Urbana Open Data Portal")