Code Editor for app.py

import streamlit as st
import pandas as pd
import plotly.express as px
from us_state_abbrev import us_state_abbrev

# DATA SOURCE: https://www.bbc.com/news/articles/cvglg3klrpzo

def standardize_state_names(state):
    """Standardize state names to full names."""
    # First, check if it's already a full name
    abbrev = us_state_abbrev[state]
    return abbrev

@st.cache_data
def load_data():
    df = pd.read_csv('./result.csv')
    df['Candidate'] = df['Candidate'].str.replace(' wins', '', regex=False)
    df['Party'] = df['Party'].str.replace(' (incumbent)', '', regex=False)
    df['Votes'] = df['Votes'].str.replace(',', '').astype(int)
    df['Electoral votes'] = df['Electoral votes'].astype(int)

    # Calculate voting power
    state_data = df.groupby('State').agg({
        'Votes': 'sum',
        'Electoral votes': 'sum'
    }).reset_index()
    state_data.loc[state_data['State'] == 'Nevada', 'Electoral votes'] = 6
    state_data.loc[state_data['State'] == 'Arizona', 'Electoral votes'] = 11
    state_data['Voting Power'] = (state_data['Electoral votes'] / state_data['Votes']) * 1000000  # Multiply by 1M for readability
    
    # Get winners and reset index to ensure proper alignment
    winners = df.loc[df.groupby('State')['Votes'].idxmax()].reset_index(drop=True)

    # Merge winners into state_data
    state_data = state_data.merge(
        winners[['State', 'Party']],
        on='State',
        how='left'
    )

    # Rename Party column to Winner
    state_data = state_data.rename(columns={'Party': 'Winner'})
    
    return df, state_data

def percentage_to_float(value):
    return float(value.strip('%')) / 100 if isinstance(value, str) else value

def create_vote_share_map(state_data, filter_type, selected_filter, color_scheme):
    fig = px.choropleth(
        state_data,
        locations='State_abv',
        locationmode='USA-states',
        color='Vote share',
        hover_name='State',
        hover_data={'Vote share': ':.1%', 'Votes': ':,', 'Electoral votes': True},
        color_continuous_scale=color_scheme,
        scope='usa',
        labels={'Vote share': 'Vote Share'}
    )
    
    fig.update_coloraxes(colorbar_tickformat='.0%')
    fig.update_layout(
        title_text=f'Election Results - {filter_type}: {selected_filter}',
        geo_scope='usa',
    )
    return fig

def create_voting_power_map(state_data):
    fig = px.choropleth(
        state_data,
        locations='State_abv',
        locationmode='USA-states',
        color='Voting Power',
        hover_name='State',
        hover_data={
            'Voting Power': ':.4f',
            'Electoral votes': True,
            'Votes': ':,',
        },
        color_continuous_scale='Viridis',
        scope='usa',
        labels={'Voting Power': 'Electoral Votes per Million Voters'}
    )
    
    fig.update_layout(
        title_text='State Voting Power (Electoral Votes per Million Voters)',
        geo_scope='usa',
    )
    return fig

def main():
    st.title('US Election Results')

    # Load the data
    df, state_data = load_data()

    # Get unique parties and candidates
    parties = df['Party'].unique()
    candidates = df['Candidate'].unique()

    # Create sidebar for user input
    st.sidebar.header('Filter Options')
    
    # Add visualization type selector
    viz_type = st.sidebar.radio('Visualization:', ('Vote Share', 'Voting Power'))
    
    if viz_type == 'Vote Share':
        filter_type = st.sidebar.radio('Filter by:', ('Party', 'Candidate'))
        
        if filter_type == 'Party':
            selected_filter = st.sidebar.selectbox('Select a party:', parties)
            filtered_df = df[df['Party'] == selected_filter]
        else:
            selected_filter = st.sidebar.selectbox('Select a candidate:', candidates)
            filtered_df = df[df['Candidate'] == selected_filter]

        # Process vote share data
        filtered_df['Vote share'] = filtered_df['Vote share'].apply(percentage_to_float)
        state_data = filtered_df.groupby('State').agg({
            'Votes': 'sum',
            'Vote share': 'mean',
            'Electoral votes': 'sum'
        }).reset_index()

        # Standardize state names
        state_data['State_abv'] = state_data['State'].apply(standardize_state_names)
        state_data['Vote share %'] = state_data['Vote share'].apply(lambda x: f"{x:.1%}")

        # Determine color scheme
        color = "Greys"
        if filter_type == 'Party':
            if selected_filter == "Republican":
                color = "Reds"
            if selected_filter == "Democrat":
                color = "Blues"
        else:
            if selected_filter == "Donald Trump":
                color = "Reds"
            if selected_filter == "Kamala Harris":
                color = "Blues"

        # Create and display vote share map
        fig = create_vote_share_map(state_data, filter_type, selected_filter, color)
        st.plotly_chart(fig)
        
    else:  # Voting Power visualization
        # Prepare state data for voting power visualization
        state_data['State_abv'] = state_data['State'].apply(standardize_state_names)
        
        # Create and display voting power map
        fig = create_voting_power_map(state_data)
        st.plotly_chart(fig)
        
        # Add explanation of voting power
        st.info("""
        **Understanding Voting Power**
        
        The voting power metric shows how much electoral influence each voter has in different states. 
        It is calculated as: (Electoral Votes / Total Votes) × 1,000,000
        
        - Higher values (lighter colors) indicate states where each voter has more electoral influence
        - Lower values (darker colors) indicate states where each voter has less electoral influence
        """)

    # Display the data table
    st.subheader('Detailed Results')
    display_cols = ['State', 'Votes', 'Electoral votes']
    if viz_type == 'Vote Share':
        display_cols.append('Vote share %')
    st.dataframe(state_data[display_cols])

if __name__ == '__main__':
    main()