app.py

import streamlit as st
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
from datetime import date, timedelta
import random



# [All the scheduling functions and analytics functions here]



# 1. create_schedule
def create_schedule(num_teams, num_conferences, num_inter_games):
    full_schedule = []
    for i in range(num_conferences):
        conference_name = chr(65 + i)  # 'A', 'B', 'C', 'D', ...
        combined_schedule = combine_schedules(conference_name, num_teams, num_inter_games)
        assigned_dates = assign_dates_to_matches(combined_schedule)
        full_schedule.extend(assigned_dates)
    return pd.DataFrame(full_schedule, columns=["Team 1", "Team 2", "Date"])

# 2. combine_schedules
def combine_schedules(conference_name, num_teams, num_inter_games):
    intra_conf_matches = generate_intra_conference_schedule(conference_name, num_teams)
    inter_conf_matches = generate_inter_conference_schedule(conference_name, num_teams, num_inter_games)
    return intra_conf_matches + inter_conf_matches

# 3. generate_intra_conference_schedule
def generate_intra_conference_schedule(conference_name, num_teams):
    teams = [f"{conference_name}{i}" for i in range(1, num_teams + 1)]
    matches = []
    for i in range(len(teams)):
        for j in range(i+1, len(teams)):
            matches.append((teams[i], teams[j]))
            matches.append((teams[j], teams[i]))  # Home and away
    return matches

# 4. generate_inter_conference_schedule
def generate_inter_conference_schedule(conference_name, num_teams, num_inter_games):
    current_conf_teams = [f"{conference_name}{i}" for i in range(1, num_teams + 1)]
    other_confs = [chr(65 + i) for i in range(4) if chr(65 + i) != conference_name]
    other_conf_teams = [f"{conf}{i}" for conf in other_confs for i in range(1, num_teams + 1)]
    matches = []
    for team in current_conf_teams:
        opponents = random.sample(other_conf_teams, num_inter_games)
        for opp in opponents:
            matches.append((team, opp))
    return matches

# 5. assign_dates_to_matches
def assign_dates_to_matches(matches):
    start_date = date(2022, 11, 6)
    end_date = date(2023, 3, 1)
    available_dates = [start_date + timedelta(days=i) for i in range((end_date - start_date).days) if (start_date + timedelta(days=i)).weekday() in [0, 2, 3, 5]]
    random.shuffle(available_dates)
    
    # Ensure cyclic reuse of dates
    extended_dates = available_dates * (len(matches) // len(available_dates) + 1)
    
    return [(match[0], match[1], extended_dates[i]) for i, match in enumerate(matches)]


# 6. generate_mock_historical_data
def generate_mock_historical_data(num_teams, num_conferences, num_inter_games, start_date, end_date):
    full_schedule = []
    for i in range(num_conferences):
        conference_name = chr(65 + i)
        combined_schedule = combine_schedules(conference_name, num_teams, num_inter_games)
        shuffled_dates = assign_dates_to_matches(combined_schedule)
        random.shuffle(shuffled_dates)
        for match in shuffled_dates:
            full_schedule.append({
                "Team 1": match[0],
                "Team 2": match[1],
                "Date": match[2]
            })
    return pd.DataFrame(full_schedule)


# Team Workload Analysis
def team_workload_analysis(schedule_df):
    # Check if the DataFrame is None
    if schedule_df is None:
        plt.figure(figsize=(10, 6))
        plt.text(0.5, 0.5, 'Please generate the schedule first before viewing analytics.', 
                 horizontalalignment='center', verticalalignment='center', 
                 fontsize=14, color='red')
        plt.axis('off')
        plt.tight_layout()
        plt.show()
        return
    """Generate a bar chart showing the number of matches each team has per week."""
    schedule_df['Week'] = schedule_df['Date'].dt.isocalendar().week
    team_counts = schedule_df.groupby(['Week', 'Team 1']).size().unstack().fillna(0)
    
    # Plot
    team_counts.plot(kind='bar', stacked=True, figsize=(15, 7), cmap='Oranges')
    plt.title('Team Workload Analysis')
    plt.ylabel('Number of Matches')
    plt.xlabel('Week Number')
    plt.tight_layout()
    plt.legend(title='Teams', bbox_to_anchor=(1.05, 1), loc='upper left')
    plt.show()

# Match Distribution
def match_distribution(schedule_df):
    # Check if the DataFrame is None
    if schedule_df is None:
        plt.figure(figsize=(10, 6))
        plt.text(0.5, 0.5, 'Please generate the schedule first before viewing analytics.', 
                 horizontalalignment='center', verticalalignment='center', 
                 fontsize=14, color='red')
        plt.axis('off')
        plt.tight_layout()
        plt.show()
        return
    """Generate a histogram showing match distribution across months."""
    schedule_df['Month'] = schedule_df['Date'].dt.month_name()
    month_order = ['November', 'December', 'January', 'February', 'March']
    
    # Plot
    plt.figure(figsize=(10, 6))
    sns.countplot(data=schedule_df, x='Month', order=month_order, palette='Oranges_r')
    plt.title('Match Distribution Across Months')
    plt.ylabel('Number of Matches')
    plt.xlabel('Month')
    plt.tight_layout()
    plt.show()

# Inter-Conference Match Analysis
def inter_conference_analysis(schedule_df):
    # Check if the DataFrame is None
    if schedule_df is None:
        plt.figure(figsize=(10, 6))
        plt.text(0.5, 0.5, 'Please generate the schedule first before viewing analytics.', 
                 horizontalalignment='center', verticalalignment='center', 
                 fontsize=14, color='red')
        plt.axis('off')
        plt.tight_layout()
        plt.show()
        return
    """Generate a heatmap showing inter-conference match frequencies."""
    # Extract the conference from the team names
    schedule_df['Conference 1'] = schedule_df['Team 1'].str[0]
    schedule_df['Conference 2'] = schedule_df['Team 2'].str[0]
    
    # Filter out intra-conference matches
    inter_conference_df = schedule_df[schedule_df['Conference 1'] != schedule_df['Conference 2']]
    
    # Create a crosstab for the heatmap
    heatmap_data = pd.crosstab(inter_conference_df['Conference 1'], inter_conference_df['Conference 2'])
    
    # Ensure every conference combination has a value
    all_conferences = schedule_df['Conference 1'].unique()
    for conf in all_conferences:
        if conf not in heatmap_data.columns:
            heatmap_data[conf] = 0
        if conf not in heatmap_data.index:
            heatmap_data.loc[conf] = 0
    
    heatmap_data = heatmap_data.sort_index().sort_index(axis=1)
    
    # Plot
    plt.figure(figsize=(8, 6))
    sns.heatmap(heatmap_data, annot=True, cmap='Oranges', linewidths=.5, cbar_kws={'label': 'Number of Matches'})
    plt.title('Inter-Conference Match Analysis')
    plt.ylabel('Conference 1')
    plt.xlabel('Conference 2')
    plt.show()

# Commissioner Analytics
def commissioner_analytics(schedule_df, commissioners):
    # Check if the DataFrame is None
    if schedule_df is None:
        plt.figure(figsize=(10, 6))
        plt.text(0.5, 0.5, 'Please generate the schedule first before viewing analytics.', 
                 horizontalalignment='center', verticalalignment='center', 
                 fontsize=14, color='red')
        plt.axis('off')
        plt.tight_layout()
        plt.show()
        return
    """Generate a bar chart showing matches overseen by each commissioner."""
    # Assuming each commissioner oversees a specific conference
    comm_dict = {f"Conference {chr(65+i)}": comm for i, comm in enumerate(commissioners)}
    schedule_df['Commissioner'] = schedule_df['Conference 1'].map(comm_dict)
    
    # Count matches overseen by each commissioner
    commissioner_counts = schedule_df['Commissioner'].value_counts()
    
    # Plot using matplotlib
    plt.figure(figsize=(10, 6))
    plt.bar(commissioner_counts.index, commissioner_counts.values, color='orange')
    plt.title('Matches Overseen by Each Commissioner')
    plt.ylabel('Number of Matches')
    plt.xlabel('Commissioner')
    plt.xticks(rotation=45)
    plt.tight_layout()
    plt.show()



# Streamlit App

st.title("Basketball Game Schedule Generator")

if 'num_teams' not in st.session_state:
    st.session_state.num_teams = 10
if 'num_conferences' not in st.session_state:
    st.session_state.num_conferences = 4
if 'num_inter_games' not in st.session_state:
    st.session_state.num_inter_games = 3

# Initialize session state for schedule_df and st.session_state.historical_data
if 'schedule_df' not in st.session_state:
    st.session_state.schedule_df = None

if 'st.session_state.historical_data' not in st.session_state:
    st.session_state.historical_data = None

if st.session_state.historical_data is None:
    st.session_state.historical_data = generate_mock_historical_data(st.session_state.num_teams, st.session_state.num_conferences, st.session_state.num_inter_games, date(2022, 11, 6), date(2023, 3, 1))
    st.session_state.historical_data['Date'] = pd.to_datetime(st.session_state.historical_data['Date'])


    
# Configuration UI
st.header("Configuration")

st.session_state.num_teams = st.number_input("Number of teams per conference:", min_value=2, value=st.session_state.num_teams)
st.session_state.num_conferences = st.number_input("Number of conferences:", min_value=2, value=st.session_state.num_conferences)
st.session_state.num_inter_games = st.number_input("Number of inter-conference games per team:", min_value=1, value=st.session_state.num_inter_games)


commissioners = st.multiselect("Add commissioners:", options=[], default=[])

add_commissioner = st.text_input("New commissioner name:")
if add_commissioner:
    commissioners.append(add_commissioner)

# Schedule Generation
if st.button("Generate Schedule"):
    st.session_state.schedule_df = create_schedule(num_teams, num_conferences, num_inter_games)
    if st.session_state.schedule_df is not None:
        st.session_state.schedule_df['Date'] = pd.to_datetime(st.session_state.schedule_df['Date'])


# Schedule Viewing
st.header("View Schedule")
conference_selector = st.selectbox("Select conference to view schedule:", options=["All"] + [f"Conference {chr(65+i)}" for i in range(st.session_state.num_conferences)])
if st.session_state.schedule_df is not None:
    if conference_selector == "All":
        st.dataframe(st.session_state.schedule_df)
    else:
        filtered_schedule = st.session_state.schedule_df[(st.session_state.schedule_df["Team 1"].str.startswith(conference_selector)) | (st.session_state.schedule_df["Team 2"].str.startswith(conference_selector))]
        st.dataframe(filtered_schedule)

# Analytics & Comparisons
st.header("Analytics & Comparisons")
analytics_option = st.selectbox("Choose an analysis type:", ["Team Workload Analysis", "Match Distribution", "Inter-Conference Match Analysis", "Commissioner Analytics"])
st.session_state.historical_data['Date'] = pd.to_datetime(st.session_state.historical_data['Date'])

if analytics_option == "Team Workload Analysis":
    st.subheader("Historical Data")
    st.pyplot(team_workload_analysis(st.session_state.historical_data))
    st.subheader("Current Data")
    st.pyplot(team_workload_analysis(st.session_state.schedule_df))

elif analytics_option == "Match Distribution":
    st.subheader("Historical Data")
    st.pyplot(match_distribution(st.session_state.historical_data))
    st.subheader("Current Data")
    st.pyplot(match_distribution(st.session_state.schedule_df))

elif analytics_option == "Inter-Conference Match Analysis":
    st.subheader("Historical Data")
    st.pyplot(inter_conference_analysis(st.session_state.historical_data))
    st.subheader("Current Data")
    st.pyplot(inter_conference_analysis(st.session_state.schedule_df))

elif analytics_option == "Commissioner Analytics":
    st.subheader("Historical Data")
    st.pyplot(commissioner_analytics(st.session_state.historical_data, commissioners))
    st.subheader("Current Data")
    st.pyplot(commissioner_analytics(st.session_state.schedule_df, commissioners))
else:
    st.warning("Please generate the schedule first before viewing analytics.")

# Export functionality can be added later