app.py

import numpy as np
import pandas as pd
import streamlit as st
import gspread
import plotly.figure_factory as ff
from itertools import combinations

scope = ['https://www.googleapis.com/auth/spreadsheets',
          "https://www.googleapis.com/auth/drive"]

credentials = {
  "type": "service_account",
  "project_id": "sheets-api-connect-378620",
  "private_key_id": "1005124050c80d085e2c5b344345715978dd9cc9",
  "private_key": "-----BEGIN PRIVATE KEY-----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\nJuElxA33rFEjbgbzdyrE768=\n-----END PRIVATE KEY-----\n",
  "client_email": "gspread-connection@sheets-api-connect-378620.iam.gserviceaccount.com",
  "client_id": "106625872877651920064",
  "auth_uri": "https://accounts.google.com/o/oauth2/auth",
  "token_uri": "https://oauth2.googleapis.com/token",
  "auth_provider_x509_cert_url": "https://www.googleapis.com/oauth2/v1/certs",
  "client_x509_cert_url": "https://www.googleapis.com/robot/v1/metadata/x509/gspread-connection%40sheets-api-connect-378620.iam.gserviceaccount.com"
}

gc = gspread.service_account_from_dict(credentials)

st.set_page_config(layout="wide")

master_hold = 'https://docs.google.com/spreadsheets/d/15flX6E7lPxu_HC7IOHpB3VEg2Am1AmtxTo9c2y_I-Mw/edit?gid=676575006#gid=676575006'

@st.cache_resource(ttl = 301)
def init_baselines():
          sh = gc.open_by_url(master_hold)
          worksheet = sh.worksheet('ADPs (model)')
          adp_hold = pd.DataFrame(worksheet.get_all_records())
          adp_hold = adp_hold[['Player', 'Team', 'Bye', 'Position', 'Position Rank', 'Underdog', 'MFL10', 'Draftkings', 'AVG', 'Projection', 'Proj ADP', 'Diff']]
          adp_table = adp_hold.drop_duplicates(subset='Player')
          adp_table.replace('', np.nan, inplace=True)
          
          worksheet = sh.worksheet('Stacks (model)')
          stacks_hold = pd.DataFrame(worksheet.get_all_records())
          stacks_table = stacks_hold.drop_duplicates(subset='Team')
          
          worksheet = sh.worksheet('Player Level Projections')
          proj_hold = pd.DataFrame(worksheet.get_all_records())
          proj_table = proj_hold[['Player', 'Team', 'Pos', 'Pass Yards', 'PassTD', 'Rush Yards', 'RushTD', 'Receptions', 'Rec Yards', 'RecTD', 'Proj']]
          proj_table.replace('', np.nan, inplace=True)
          
          return adp_table, stacks_table, proj_table

adp_table, stacks_table, proj_table = init_baselines()

tab1, tab2, tab3, tab4 = st.tabs(["ADPs and Ranks", "Team Projections", 'Player Projections', "Stack Finder"])

def convert_df_to_csv(df):
    return df.to_csv().encode('utf-8')

with tab1:
    col1, col2 = st.columns([1, 5])
    
    with col1:
        if st.button("Load/Reset Data", key='reset1'):
              st.cache_data.clear()
              adp_table, stacks_table, proj_table = init_baselines()
        site_var1 = st.radio("What site are you playing?", ('Underdog', 'Draftkings'), key='site_var1')
        split_var1 = st.radio("Would you like to run stack analysis for the full slate or individual teams?", ('All Teams', 'Specific Teams'), key='split_var1')
        if split_var1 == 'Specific Teams':
            team_var1 = st.multiselect('Which teams would you like to include in the analysis?', options = adp_table['Team'].unique(), key='team_var1')
        elif split_var1 == 'All Teams':
            team_var1 = adp_table.Team.unique().tolist()
        pos_split1 = st.radio("Are you viewing all positions, specific groups, or specific positions?", ('All Positions', 'Specific Positions'), key='pos_split1')
        if pos_split1 == 'Specific Positions':
            pos_var1 = st.multiselect('What Positions would you like to view?', options = ['QB', 'RB', 'WR', 'TE'])
        elif pos_split1 == 'All Positions':
            pos_var1 = adp_table.Position.unique().tolist()
    
    with col2:
        stack_hold_container = st.empty()
        working_baselines = adp_table.copy()
        
        if pos_split1 == 'All Positions':
            raw_baselines = working_baselines
        elif pos_split1 != 'All Positions':
            raw_baselines = working_baselines[working_baselines['Position'].str.contains('|'.join(pos_var1))]

        if split_var1 == 'All Teams':
            raw_baselines = raw_baselines
        elif split_var1 != 'All Teams':
            raw_baselines = raw_baselines[raw_baselines['Team'].str.contains('|'.join(team_var1))]
            
        display_frame = raw_baselines.copy()
        display_frame = display_frame.sort_values(by='Proj ADP', ascending=True)
        
        with stack_hold_container:
            stack_hold_container = st.empty()
            st.dataframe(display_frame.style.background_gradient(axis=0).background_gradient(cmap='RdYlGn_r').background_gradient(cmap='RdYlGn', subset=['Projection', 'Diff']).format(precision=2), height=750, use_container_width = True)
        st.download_button(
                label="Export Tables",
                data=convert_df_to_csv(display_frame),
                file_name='NFL_ADPs_Ranks_export.csv',
                mime='text/csv',
        )
        
with tab2:
    if st.button("Load/Reset Data", key='reset2'):
          st.cache_data.clear()
          adp_table, stacks_table, proj_table = init_baselines()
    stack_hold_container = st.empty()
    working_baselines = stacks_table.copy()
    display_frame = working_baselines.copy()
    
    with stack_hold_container:
        stack_hold_container = st.empty()
        st.dataframe(display_frame.style.background_gradient(axis=0).background_gradient(cmap='RdYlGn').background_gradient(cmap='RdYlGn_r', subset=['Min ADP', 'Avg ADP', 'Max ADP', 'Value Build', 'Value']).format(precision=2), height=750, use_container_width = True)
    st.download_button(
            label="Export Tables",
            data=convert_df_to_csv(display_frame),
            file_name='NFL_Team_Projections_export.csv',
            mime='text/csv',
    )
    
with tab3:
    if st.button("Load/Reset Data", key='reset3'):
          st.cache_data.clear()
          adp_table, stacks_table, proj_table = init_baselines()
    stack_hold_container = st.empty()
    working_baselines = proj_table.copy()
    display_frame = working_baselines.copy()
    
    with stack_hold_container:
        stack_hold_container = st.empty()
        st.dataframe(display_frame.style.background_gradient(axis=0).background_gradient(cmap='RdYlGn').format(precision=2), height=750, use_container_width = True)
    st.download_button(
            label="Export Tables",
            data=convert_df_to_csv(display_frame),
            file_name='NFL_Player_Projections_export.csv',
            mime='text/csv',
    )
    
with tab4:
    col1, col2 = st.columns([1, 5])
    
    with col1:
        if st.button("Load/Reset Data", key='reset4'):
              st.cache_data.clear()
              adp_table, stacks_table, proj_table = init_baselines()
        site_var4 = st.radio("What site are you playing?", ('Underdog', 'Draftkings'), key='site_var2')
        split_var4 = st.radio("Would you like to run stack analysis for the full slate or individual teams?", ('All Teams', 'Specific Teams'), key='split_var4')
        if split_var4 == 'Specific Teams':
            team_var4 = st.multiselect('Which teams would you like to include in the analysis?', options = adp_table['Team'].unique(), key='team_var4')
        elif split_var4 == 'All Teams':
            team_var4 = adp_table.Team.unique().tolist()
        pos_split4 = st.radio("Are you viewing all positions, specific groups, or specific positions?", ('All Positions', 'Specific Positions'), key='pos_split4')
        if pos_split4 == 'Specific Positions':
            pos_var4 = st.multiselect('What Positions would you like to view?', options = ['QB', 'RB', 'WR', 'TE'], key='pos_var4')
        elif pos_split4 == 'All Positions':
            pos_var4 = adp_table.Position.unique().tolist()
        if site_var4 == 'Underdog':
            adp_dict = dict(zip(adp_table.Player, adp_table.Underdog))
        elif site_var4 == 'Draftkings':
           adp_dict = dict(zip(adp_table.Player, adp_table.Draftkings))
        size_var4 = st.number_input('What size of stacks are you analyzing?', min_value = 3, max_value = 6, step=1)
        stack_size = size_var4
        cut_var4 = st.radio("Do you want to remove stacks with a negative average value?", ('Yes', 'No'), key='cut_var4')
        if cut_var4 == "Yes":
            cut_sequence = 1
        elif cut_var4 == "No":
            cut_sequence = 0
    
        team_dict = dict(zip(adp_table.Player, adp_table.Team))
        proj_dict = dict(zip(adp_table.Player, adp_table.Projection))
        diff_dict = dict(zip(adp_table.Player, adp_table.Diff))
    
    with col2:
        stack_hold_container = st.empty()
        if st.button('Run stack analysis'):
            comb_list = []
            if pos_split4 == 'All Positions':
                raw_baselines = adp_table.copy()
            elif pos_split4 != 'All Positions':
                raw_baselines = adp_table[adp_table['Position'].str.contains('|'.join(pos_var4))]
    
            for cur_team in team_var4:
                working_baselines = raw_baselines.copy()
                working_baselines = working_baselines[working_baselines['Team'] == cur_team]
                order_list = working_baselines['Player']
    
                comb = combinations(order_list, stack_size)
    
                for i in list(comb):
                    comb_list.append(i)
    
            comb_DF = pd.DataFrame(comb_list)
    
            if stack_size == 3:
                comb_DF['Team'] = comb_DF[0].map(team_dict)
    
                comb_DF['Proj'] = comb_DF.apply(lambda row: pd.Series([proj_dict.get(row[i], None) for i in range(3)]).sum(), axis=1)
    
                comb_DF['ADP_1'] = comb_DF[0].map(adp_dict)
                comb_DF['ADP_2'] = comb_DF[1].map(adp_dict)
                comb_DF['ADP_3'] = comb_DF[2].map(adp_dict)
    
                comb_DF['Value'] = comb_DF.apply(lambda row: pd.Series([diff_dict.get(row[i], None) for i in range(3)]).mean(), axis=1)
                
            elif stack_size == 4:
                comb_DF['Team'] = comb_DF[0].map(team_dict)
    
                comb_DF['Proj'] = comb_DF.apply(lambda row: pd.Series([proj_dict.get(row[i], None) for i in range(4)]).sum(), axis=1)
    
                comb_DF['ADP_1'] = comb_DF[0].map(adp_dict)
                comb_DF['ADP_2'] = comb_DF[1].map(adp_dict)
                comb_DF['ADP_3'] = comb_DF[2].map(adp_dict)
                comb_DF['ADP_4'] = comb_DF[3].map(adp_dict)
    
                comb_DF['Value'] = comb_DF.apply(lambda row: pd.Series([diff_dict.get(row[i], None) for i in range(4)]).mean(), axis=1)
                
            elif stack_size == 5:
                comb_DF['Team'] = comb_DF[0].map(team_dict)
    
                comb_DF['Proj'] = comb_DF.apply(lambda row: pd.Series([proj_dict.get(row[i], None) for i in range(5)]).sum(), axis=1)
    
                comb_DF['ADP_1'] = comb_DF[0].map(adp_dict)
                comb_DF['ADP_2'] = comb_DF[1].map(adp_dict)
                comb_DF['ADP_3'] = comb_DF[2].map(adp_dict)
                comb_DF['ADP_4'] = comb_DF[3].map(adp_dict)
                comb_DF['ADP_5'] = comb_DF[4].map(adp_dict)
    
                comb_DF['Value'] = comb_DF.apply(lambda row: pd.Series([diff_dict.get(row[i], None) for i in range(5)]).mean(), axis=1)
            
            elif stack_size == 6:
                comb_DF['Team'] = comb_DF[0].map(team_dict)
    
                comb_DF['Proj'] = comb_DF.apply(lambda row: pd.Series([proj_dict.get(row[i], None) for i in range(6)]).sum(), axis=1)
    
                comb_DF['ADP_1'] = comb_DF[0].map(adp_dict)
                comb_DF['ADP_2'] = comb_DF[1].map(adp_dict)
                comb_DF['ADP_3'] = comb_DF[2].map(adp_dict)
                comb_DF['ADP_4'] = comb_DF[3].map(adp_dict)
                comb_DF['ADP_5'] = comb_DF[4].map(adp_dict)
                comb_DF['ADP_6'] = comb_DF[5].map(adp_dict)
                
                comb_DF['Value'] = comb_DF.apply(lambda row: pd.Series([diff_dict.get(row[i], None) for i in range(6)]).mean(), axis=1)
    
            comb_DF = comb_DF.sort_values(by='Proj', ascending=False)
            
            if cut_sequence == 1:
                cut_var = 0
        
                if stack_size == 3:
                    while cut_var <= int(len(comb_DF)):
                        try:
                            if int(cut_var) == 0:
                                cur_proj = float(comb_DF.iat[cut_var,4])
                                cur_own = 0
                            elif int(cut_var) >= 1:
                                check_own = float(comb_DF.iat[cut_var,8])
                                if check_own < cur_own:
                                    comb_DF = comb_DF.drop([cut_var])
                                    cur_own = cur_own
                                    cut_var = cut_var - 1
                                    comb_DF = comb_DF.reset_index()
                                    comb_DF = comb_DF.drop(['index'], axis=1)
                                elif check_own >= cur_own:
                                    cur_own = float(comb_DF.iat[cut_var,8])
                                    cut_var = cut_var
                            cut_var += 1
                        except:
                            cut_var += 1
                            
                elif stack_size == 4:
                    while cut_var <= int(len(comb_DF)):
                        try:
                            if int(cut_var) == 0:
                                cur_proj = float(comb_DF.iat[cut_var,5])
                                cur_own = 0
                            elif int(cut_var) >= 1:
                                check_own = float(comb_DF.iat[cut_var,10])
                                if check_own < cur_own:
                                    comb_DF = comb_DF.drop([cut_var])
                                    cur_own = cur_own
                                    cut_var = cut_var - 1
                                    comb_DF = comb_DF.reset_index()
                                    comb_DF = comb_DF.drop(['index'], axis=1)
                                elif check_own >= cur_own:
                                    cur_own = float(comb_DF.iat[cut_var,10])
                                    cut_var = cut_var
                            cut_var += 1
                        except:
                            cut_var += 1
                elif stack_size == 5:
                    while cut_var <= int(len(comb_DF)):
                        try:
                            if int(cut_var) == 0:
                                cur_proj = float(comb_DF.iat[cut_var,6])
                                cur_own = 0
                            elif int(cut_var) >= 1:
                                check_own = float(comb_DF.iat[cut_var,12])
                                if check_own < cur_own:
                                    comb_DF = comb_DF.drop([cut_var])
                                    cur_own = cur_own
                                    cut_var = cut_var - 1
                                    comb_DF = comb_DF.reset_index()
                                    comb_DF = comb_DF.drop(['index'], axis=1)
                                elif check_own >= cur_own:
                                    cur_own = float(comb_DF.iat[cut_var,12])
                                    cut_var = cut_var
                            cut_var += 1
                        except:
                            cut_var += 1
                elif stack_size == 6:
                    while cut_var <= int(len(comb_DF)):
                        try:
                            if int(cut_var) == 0:
                                cur_proj = float(comb_DF.iat[cut_var,7])
                                cur_own = 0
                            elif int(cut_var) >= 1:
                                check_own = float(comb_DF.iat[cut_var,14])
                                if check_own < cur_own:
                                    comb_DF = comb_DF.drop([cut_var])
                                    cur_own = cur_own
                                    cut_var = cut_var - 1
                                    comb_DF = comb_DF.reset_index()
                                    comb_DF = comb_DF.drop(['index'], axis=1)
                                elif check_own >= cur_own:
                                    cur_own = float(comb_DF.iat[cut_var,14])
                                    cut_var = cut_var
                            cut_var += 1
                        except:
                            cut_var += 1
    
            with stack_hold_container:
                stack_hold_container = st.empty()
                st.dataframe(comb_DF.style.background_gradient(axis=0).background_gradient(cmap='RdYlGn').format(precision=2), use_container_width = True)
            st.download_button(
                    label="Export Tables",
                    data=convert_df_to_csv(comb_DF),
                    file_name='NFL_Stack_Options_export.csv',
                    mime='text/csv',
            )