src/streamlit_app.py

import streamlit as st
import numpy as np
import pandas as pd
import pymongo
import os
import time

st.set_page_config(layout="wide")

@st.cache_resource
def init_conn():
        uri = os.getenv('mongo_uri')
        client = pymongo.MongoClient(uri, retryWrites=True, serverSelectionTimeoutMS=500000)
        db = client["MLB_Database"]

        return db
    
db = init_conn()

st.markdown("""
<style>
    /* Tab styling */
    .stTabs [data-baseweb="tab-list"] {
        gap: 8px;
        padding: 4px;
    }
    .stTabs [data-baseweb="tab"] {
        height: 50px;
        white-space: pre-wrap;
        background-color: #DAA520;
        color: white;
        border-radius: 10px;
        gap: 1px;
        padding: 10px 20px;
        font-weight: bold;
        transition: all 0.3s ease;
    }
    .stTabs [aria-selected="true"] {
        background-color: #DAA520;
        border: 3px solid #FFD700;
        color: white;
    }
    .stTabs [data-baseweb="tab"]:hover {
        background-color: #FFD700;
        cursor: pointer;
    }
    div[data-baseweb="select"] > div {
        background-color: #DAA520;
        color: white;
    }
    div{
        box-sizing: content-box !important;
    }
</style>""", unsafe_allow_html=True)

def paginate_dataframe(df, page_size):
    total_rows = len(df)
    total_pages = (total_rows + page_size -1) // page_size

    if 'current_page' not in st.session_state:
        st.session_state['current_page'] = 0
    
    start_idx = st.session_state['current_page'] * page_size
    end_idx = start_idx + page_size

    current_page_data = df.iloc[start_idx:end_idx]

    return current_page_data, total_pages, total_rows

def display_paginated_table(df, page_size):
    current_page_data, total_pages, total_rows = paginate_dataframe(df, page_size)

    st.dataframe(current_page_data.style.background_gradient(axis=0).background_gradient(cmap='RdYlGn_r').format(precision=2), height = 500, use_container_width = True, hide_index = True)

    col1, col2, col3, col4 = st.columns(1, 2, 2, 1)
    with col1:
        if st.button('⏮️ First', disabled=st.session_state.current_page == 0):
            st.session_state.current_page = 0
            st.rerun()
    with col2:
        if st.button('⬅️ Previous', disabled=st.session_state.current_page == 0):
            st.session_state.current_page -= 1
            st.rerun()
    with col3:
        if st.button('Next ➡️', disabled=st.session_state.current_page >= total_pages - 1):
            st.session_state.current_page += 1
            st.rerun()
    with col4:
        if st.button('Last ⏭️', disabled=st.session_state.current_page >= total_pages - 1):
            st.session_state.current_page = total_pages - 1
            st.rerun()

@st.cache_resource(ttl = 61)
def init_baselines():

    db_pulls = ['Bullpen_Data', 'Hitter_Agg_Merge', 'Hitter_Long_Merge', 'Hitter_Short_Merge', 'Pitcher_Agg_Merge', 'Pitcher_Long_Merge', 'Pitcher_Short_Merge',
                'Slate_Hitters_Merge', 'Slate_Teams_Merge', 'Starting_Pitchers', 'True_AVG_Split', 'Pitcher_Info', 'Hitter_Info']

    for table in db_pulls:
        collection = db[table] 
        cursor = collection.find()
        df = pd.DataFrame(cursor)

        if table == 'Bullpen_Data':
            bp_data = df.drop(columns = ['_id'])
        elif table == 'Hitter_Agg_Merge':
            hitter_agg = df.drop(columns = ['_id'])
        elif table == 'Hitter_Long_Merge':
            hitter_long = df.drop(columns = ['_id'])
        elif table == 'Hitter_Short_Merge':
            hitter_short = df.drop(columns = ['_id'])
        elif table == 'Pitcher_Agg_Merge':
            pitcher_agg = df.drop(columns = ['_id'])
        elif table == 'Pitcher_Long_Merge':
            pitcher_long = df.drop(columns = ['_id'])
        elif table == 'Pitcher_Short_Merge':
            pitcher_short = df.drop(columns = ['_id'])
        elif table == 'Slate_Hitters_Merge':
            slate_hitters = df.drop(columns = ['_id'])
        elif table == 'Slate_Teams_Merge':
            slate_team = df.drop(columns = ['_id'])
        elif table == 'Starting_Pitchers':  
            starting_pitchers = df.drop(columns = ['_id'])
        elif table == 'True_AVG_Split':
            true_avg_split = df.drop(columns = ['_id'])
        elif table == 'Pitcher_Info':
            pitcher_info = df.drop(columns = ['_id'])
        elif table == 'Hitter_Info':    
            hitter_info = df.drop(columns = ['_id'])

    return bp_data, hitter_agg, hitter_long, hitter_short, pitcher_agg, pitcher_long, pitcher_short, slate_hitters, slate_team, starting_pitchers, true_avg_split, pitcher_info, hitter_info

bp_data, hitter_agg, hitter_long, hitter_short, pitcher_agg, pitcher_long, pitcher_short, slate_hitters, slate_team, starting_pitchers, true_avg_split, pitcher_info, hitter_info = init_baselines()

pitcher_tab, hitter_tab, team_tab = st.tabs(['Pitchers', 'Hitters', 'Team'])

with pitcher_tab:
    with st.container(border = True):
        st.info('Note: Splits options are available for all baseline tables, they do not apply to True AVG, HWSr, or the Overview tables')
        col1, col2, col3, col4, col5 = st.columns(5)
        with col1:
            site_var_sp = st.selectbox('Site', ['DraftKings', 'FanDuel'], key = 'site_var_sp')
        with col2:
            table_var_sp = st.selectbox('Table', ['True AVG Splits', 'HWSr Splits', 'Current Slate Overview', 'Active Baselines', 'League Aggregate Baselines', 'League Short Term Baselines', 'League Long Term Baselines'], key = 'table_var_sp')
        with col3:
            splits_var_sp = st.selectbox('Splits', ['RHH', 'LHH', 'Overall'], key = 'splits_var_sp')
        with col4:
            team_type_sp = st.selectbox('Do you want to view all teams or Specific ones?', ['All', 'Specific'], key = 'team_type_sp')
        with col5:
            if team_type_sp == 'Specific':
                team_var_sp = st.multiselect('Select Teams', starting_pitchers['Team'].unique(), key = 'team_var_sp')
            else:
                team_var_sp = None
                st.write('All teams selected')
    
    if table_var_sp == 'True AVG Splits':
        disp_raw = true_avg_split

        if team_var_sp is not None:
            disp_raw = disp_raw[disp_raw['Team'].isin(team_var_sp)]

        disp_raw = disp_raw[['Player', 'Handedness', 'Team', 'Opp', 'Opp LHH', 'Opp RHH', 'True AVG (LHH)', 'True AVG (RHH)', 'True AVG (Overall)', 'Weighted True AVG']]

    elif table_var_sp == 'HWSr Splits':
        disp_raw = true_avg_split

        if team_var_sp is not None:
            disp_raw = disp_raw[disp_raw['Team'].isin(team_var_sp)]

        disp_raw = disp_raw[['Player', 'Handedness', 'Team', 'Opp', 'Opp LHH', 'Opp RHH', 'HWSr (LHH)', 'HWSr (RHH)', 'HWSr (Overall)', 'Weighted HWSr']]
        

    elif table_var_sp == 'Current Slate Overview':
        disp_raw = starting_pitchers

        disp_raw = disp_raw.drop(columns = ['Own Adj', 'Perf_Adj'])

        if team_var_sp is not None:
            disp_raw = disp_raw[disp_raw['Team'].isin(team_var_sp)]

    elif table_var_sp == 'Active Baselines':
        disp_raw = pitcher_info

        if splits_var_sp != 'Overall':
            disp_raw = disp_raw[disp_raw['Set'] == splits_var_sp]
        else:
            disp_raw = disp_raw[disp_raw['Set'] == 'RHH']
        
        if team_var_sp is not None:
            disp_raw = disp_raw[disp_raw['Team'].isin(team_var_sp)]

        disp_raw = disp_raw[['Names', 'DK_Salary', 'FD_Salary', 'Team', 'Opp', 'Opp_TT', 'Hand', 'K%', 'BB%', 'True AVG', 'xSLG', 'xBA', 'Hits', 'xHRs', 'xHR/PA']]
        positive_set = ['K%']

    elif table_var_sp == 'League Aggregate Baselines':
        disp_raw = pitcher_agg
        disp_raw = disp_raw[disp_raw['Set'] == splits_var_sp]
        disp_raw = disp_raw[['Player', 'PA', 'Hits', 'Singles', 'Doubles', 'Homeruns', 'Strikeoutper', 'Strikeouts', 'Walkper', 'Walks', 'xBA', 'xSLG', 'xwOBA', 'BABIP', 'AVG', 'FB%', 'True_AVG', 'xHits', 'xHRs', 'xHR/PA', 'HWSr']]

    elif table_var_sp == 'League Short Term Baselines':
        disp_raw = pitcher_short
        disp_raw = disp_raw[disp_raw['Set'] == splits_var_sp]
        disp_raw = disp_raw[['Player', 'PA', 'Hits', 'Singles', 'Doubles', 'Homeruns', 'Strikeoutper', 'Strikeouts', 'Walkper', 'Walks', 'xBA', 'xSLG', 'xwOBA', 'BABIP', 'AVG', 'FB%', 'True_AVG', 'xHits', 'xHRs', 'xHR/PA', 'HWSr']]

    elif table_var_sp == 'League Long Term Baselines':
        disp_raw = pitcher_long
        disp_raw = disp_raw[disp_raw['Set'] == splits_var_sp]
        disp_raw = disp_raw[['Player', 'PA', 'Hits', 'Singles', 'Doubles', 'Homeruns', 'Strikeoutper', 'Strikeouts', 'Walkper', 'Walks', 'xBA', 'xSLG', 'xwOBA', 'BABIP', 'AVG', 'FB%', 'True_AVG', 'xHits', 'xHRs', 'xHR/PA', 'HWSr']]

    st.session_state['sp_disp_frame'] = disp_raw
    page_var = len(st.session_state['sp_disp_frame']) / 2
    sp_disp_container = st.container(border = True)
    sp_disp_container = sp_disp_container.empty()
    
    if table_var_sp in (['League Aggregate Baselines', 'League Short Term Baselines', 'League Long Term Baselines']):

        with st.spinner("Full league baselines can take some time to load"):
            time.sleep(5)
        display_paginated_table(st.session_state['sp_disp_frame'], 50)
    else:
        with sp_disp_container:
            st.dataframe(st.session_state['sp_disp_frame'].style.background_gradient(axis=0).background_gradient(cmap='RdYlGn_r').format(precision=2), height = 500, use_container_width = True, hide_index = True)

with hitter_tab:
    with st.container(border = True):
        st.info('Note: Splits options are available for all baseline tables')
        col1, col2, col3, col4, col5 = st.columns(5)
        with col1:
            site_var_hitter = st.selectbox('Site', ['Draftkings', 'Fanduel'], key = 'site_var_hitter')
        with col2:
            table_var_hitter = st.selectbox('Table', ['Current Slate Overview', 'Active Baselines', 'League Aggregate Baselines', 'League Short Term Baselines', 'League Long Term Baselines'], key = 'table_var_hitter')
        with col3:
            splits_var_hitter = st.selectbox('Splits', ['Overall', 'RHP', 'LHP'], key = 'splits_var_hitter')
        with col4:
            position_type_hitter = st.selectbox('Do you want to view all Positions or Specific ones?', ['All', 'Specific'], key = 'position_type_hitter')
            if position_type_hitter == 'Specific':
                position_var_hitter = st.multiselect('Positions', ['C', '1B', '2B', '3B', 'SS', 'OF'], key = 'position_var_hitter')
            else:
                position_var_hitter = None
                st.write('All Positions selected')
        with col5:
            team_type_hitter = st.selectbox('Do you want to view all Teams or Specific ones?', ['All', 'Specific'], key = 'team_type_hitter')
            if team_type_hitter == 'Specific':
                team_var_hitter = st.multiselect('Select Teams', slate_hitters['Team'].unique(), key = 'team_var_hitter')
            else:
                team_var_hitter = None
                st.write('All Teams selected')

    if table_var_hitter == 'Current Slate Overview':
        disp_raw = slate_hitters
        
        if team_var_hitter is not None:
            disp_raw = disp_raw[disp_raw['Team'].isin(team_var_hitter)]

        if position_var_hitter:
            position_mask = disp_raw['Position'].apply(lambda x: any(pos in x for pos in position_var_hitter))
            disp_raw = disp_raw[position_mask]
        
        disp_raw = disp_raw[disp_raw['Set'] == site_var_hitter]

    elif table_var_hitter == 'Active Baselines':
        disp_raw = hitter_info
        
        if team_var_hitter is not None:
            disp_raw = disp_raw[disp_raw['Team'].isin(team_var_hitter)]
        
        if position_var_hitter:
            position_mask = disp_raw['Position'].apply(lambda x: any(pos in x for pos in position_var_hitter))
            disp_raw = disp_raw[position_mask]
        
        disp_raw = disp_raw.drop(columns = ['DK_SD_Salary', 'FD_SD_Salary', 'DK_Own', 'FD_Own', 'DK_player_ID', 'FD_player_ID', 'Opp_TT'])

    elif table_var_hitter == 'League Aggregate Baselines':
        disp_raw = hitter_agg
        
        if team_var_hitter is not None:
            disp_raw = disp_raw[disp_raw['Team'].isin(team_var_hitter)]

    elif table_var_hitter == 'League Short Term Baselines':
        disp_raw = hitter_short
        
        if team_var_hitter is not None:
            disp_raw = disp_raw[disp_raw['Team'].isin(team_var_hitter)]

    elif table_var_hitter == 'League Long Term Baselines':
        disp_raw = hitter_long
        
        if team_var_hitter is not None:
            disp_raw = disp_raw[disp_raw['Team'].isin(team_var_hitter)]

    st.session_state['hitter_disp_frame'] = disp_raw
    hitter_disp_container = st.container(border = True)
    hitter_disp_container = hitter_disp_container.empty()

    with hitter_disp_container:
        if table_var_hitter in (['League Aggregate Baselines', 'League Short Term Baselines', 'League Long Term Baselines']):
            with st.spinner("Full league baselines can take some time to load"):
                time.sleep(7)
            display_paginated_table(st.session_state['hitter_disp_frame'], 50)
        else:
            st.dataframe(st.session_state['hitter_disp_frame'].style.background_gradient(axis=0).background_gradient(cmap='RdYlGn').format(precision=2), height = 750, use_container_width = True, hide_index = True)

with team_tab:
    with st.container(border = True):
        col1, col2, col3 = st.columns(3)
        with col1:
            site_var_team= st.selectbox('Site', ['DraftKings', 'FanDuel'], key = 'site_var_team')
        with col2:
            table_var_team = st.selectbox('Table', ['Team Baselines', 'Bullpen Baselines'], key = 'table_var_team')
    
    if table_var_team == 'Team Baselines':
        st.session_state['team_disp_frame'] = slate_team
    elif table_var_team == 'Bullpen Baselines':
        st.session_state['team_disp_frame'] = bp_data

    team_disp_container = st.container(border = True)
    team_disp_container = team_disp_container.empty()

    with team_disp_container:
        st.dataframe(st.session_state['team_disp_frame'].style.background_gradient(axis=0).background_gradient(cmap='RdYlGn').format(precision=2), height = 750, use_container_width = True, hide_index = True)