nfs_drive_analysis.py

from functools import reduce
import warnings

import jenkspy
import pandas as pd

warnings.filterwarnings('ignore')
pd.set_option('display.max_rows', None)
pd.set_option('display.max_columns', None)
pd.set_option('display.width', None)


def time_to_float(time_str):
    minutes, seconds = map(int, time_str.split(':'))
    return float(minutes * 60 + seconds)


def clean_data(df):
    df['Start'] = df['Start'].str.replace('Own ', '')
    df['Start'] = df['Start'].astype(float)
    df['Time'] = df['Time'].apply(time_to_float)
    df = df.loc[df['Tm'] != 'League Total']
    return df


def create_bins(df, cols, classes=4):
    for col in cols:
        res = jenkspy.jenks_breaks(df[col], n_classes=classes)
        col_name = f'{col}_bin'
        df[col_name] = pd.cut(df[col], bins=res, include_lowest=True, labels=False, duplicates='drop')
        df[col_name] = df[col_name] + 1

    return df


def prep_data(df, analyze_cols):
    df = clean_data(df)
    df = create_bins(df, analyze_cols)
    return df


def process_data(path, analyze_cols):
    offense_df = pd.read_csv(path, skiprows=1)
    offense_df = prep_data(offense_df, analyze_cols)
    return offense_df

def create_analysis_dataframe(offense_data_path, defense_data_path, analyze_cols, analyze_teams):
    offense_df = process_data(offense_data_path, analyze_cols)
    defense_df = process_data(defense_data_path, analyze_cols)
    offense_df = offense_df.loc[offense_df['Tm'].isin(analyze_teams)]
    defense_df = defense_df.loc[defense_df['Tm'].isin(analyze_teams)]

    output_dfs = []
    for team in analyze_teams:
        team_df = pd.DataFrame()
        for col in analyze_cols:
            col = f'{col}_bin'
            offense_val = offense_df.loc[offense_df['Tm'] == team][col].iloc[0]
            defense_val = defense_df.loc[defense_df['Tm'] == team][col].iloc[0]
            team_col_df = pd.DataFrame({
                'stat': [col],
                f'offensive_bin_{team}': [offense_val],
                f'defensive_bin_{team}': [defense_val]
            })
            team_df = pd.concat([team_df, team_col_df], axis=0)
        team_df = team_df.reset_index(drop=True)
        output_dfs.append(team_df)
    analysis_df = reduce(lambda left, right: pd.merge(left, right, on='stat', how='inner'), output_dfs)
    return analysis_df


if __name__ == "__main__":
    # https://www.pro-football-reference.com/years/2023/
    drives_analysis_df = create_analysis_dataframe(
        offense_data_path='drive_averages.csv',
        defense_data_path='drives_against.csv',
        analyze_cols=['Sc%', 'TO%', 'Plays.1', 'Yds', 'Start', 'Time', 'Pts'],
        analyze_teams=['Kansas City Chiefs', 'San Francisco 49ers']
    )
    print(drives_analysis_df)