import pandas as pd import glob import os import plotly.graph_objects as go import json import re from datetime import datetime def analyze_game_schedules(): # 1. Define aliases for columns to handle header variations across years column_aliases = { 'Date': ['Date', 'date', 'DATE'], 'Num': ['Num', 'Number', 'game_num', 'Number of Game'], 'Day': ['Day', 'day', 'DAY'], 'Visitor': ['Visitor', 'Visiting Team', 'visitor', 'Vis', 'VISITOR'], 'Home': ['Home', 'Home Team', 'home', 'Hm', 'HOME'], 'Postponed': ['Postponed', 'postponed', 'Postponement', 'POSTPONED'], 'Makeup': ['Makeup', 'makeup', 'Makeup Date', 'MAKEUP'] } # 2. Find all CSV files in the schedules folder all_files = glob.glob("schedules/*.csv") files_to_process = [f for f in all_files if not f.endswith("2020sched-orig.csv")] df_list = [] modern_teams = [] for file_path in files_to_process: try: temp_df = pd.read_csv(file_path, low_memory=False) # Clean column names immediately temp_df.columns = [str(c).strip() for c in temp_df.columns] # Map headers to our unified names rename_map = {} for official, aliases in column_aliases.items(): for alias in aliases: if alias in temp_df.columns: rename_map[alias] = official break # Skip if we don't have the core required columns if not all(col in rename_map for col in ['Visitor', 'Home', 'Date']): continue clean_df = temp_df[list(rename_map.keys())].rename(columns=rename_map).copy() # Ensure the Date column has data clean_df = clean_df.dropna(subset=['Date']) df_list.append(clean_df) # Extract modern teams (30 franchises) from the 2025 file if "2025" in file_path: teams_2025 = set(clean_df['Visitor'].dropna().unique()) | set(clean_df['Home'].dropna().unique()) modern_teams = sorted([str(t).strip() for t in teams_2025 if len(str(t).strip()) == 3]) except Exception as e: print(f"Skipping {file_path} due to error: {e}") if not df_list: print("No data found to analyze.") return # 3. Combine into one master dataframe master_df = pd.concat(df_list, ignore_index=True, sort=False) # 4. Standardize all data for col in ['Postponed', 'Makeup', 'Date', 'Visitor', 'Home']: if col in master_df.columns: # Fill NA, convert to string, strip whitespace, remove .0 if float-converted master_df[col] = master_df[col].fillna('').astype(str).str.strip().str.replace(r'\.0$', '', regex=True) # FRANCHISE CONSOLIDATION LOGIC # Map historical codes to modern MIA for consolidation franchise_map = {'FLO': 'MIA'} if 'Visitor' in master_df.columns and 'Home' in master_df.columns: master_df['Visitor'] = master_df['Visitor'].replace(franchise_map) master_df['Home'] = master_df['Home'].replace(franchise_map) # 5. FILTERING: Remove games explicitly cancelled or not made up # Per user instructions: games = total rows minus "No makeup played" OR "Not made up" OR "Not rescheduled" cancel_mask = pd.Series(False, index=master_df.index) if 'Postponed' in master_df.columns: cancel_mask |= master_df['Postponed'].str.contains("No makeup played", case=False, na=False) if 'Makeup' in master_df.columns: cancel_mask |= master_df['Makeup'].str.contains("Not made up", case=False, na=False) if 'Makeup' in master_df.columns: cancel_mask |= master_df['Makeup'].str.contains("Not rescheduled", case=False, na=False) master_df = master_df[~cancel_mask] # 6. ROBUST DATE EXTRACTION # Instead of to_numeric, we use regex to extract the first 8 consecutive digits. def extract_8_digit_date(val): match = re.search(r'(\d{8})', val) return match.group(1) if match else "" master_df['CleanDate'] = master_df['Date'].apply(extract_8_digit_date) master_df['CleanMakeup'] = master_df['Makeup'].apply(extract_8_digit_date) # Determine final play date: Makeup date takes priority, otherwise use scheduled Date. def determine_actual_date(row): if row['CleanMakeup'] != "": return row['CleanMakeup'] return row['CleanDate'] master_df['ActualDateStr'] = master_df.apply(determine_actual_date, axis=1) # Final cleanup of the date column master_df = master_df[master_df['ActualDateStr'] != ""] master_df['Year'] = master_df['ActualDateStr'].str[:4].astype(int) master_df['MonthDayStr'] = master_df['ActualDateStr'].str[-4:] # Exclude Leap Day to keep 365-day visualization consistent master_df = master_df[master_df['MonthDayStr'] != "0229"] # 7. Map MonthDay to Day-of-Year (1-365) def get_day_of_year(mmdd): try: # Using 2023 as reference non-leap year return datetime.strptime(f"2023{mmdd}", "%Y%m%d").timetuple().tm_yday except: return None master_df['DayOfYear'] = master_df['MonthDayStr'].apply(get_day_of_year) master_df = master_df.dropna(subset=['DayOfYear']) # 8. Prepare Team Mapping for JS all_teams_global = sorted(list(set(master_df['Visitor'].unique()) | set(master_df['Home'].unique()))) all_teams_global = [t for t in all_teams_global if len(t) == 3] team_to_id = {team: i + 1 for i, team in enumerate(all_teams_global)} master_df['VisitorID'] = master_df['Visitor'].map(team_to_id).fillna(0).astype(int) master_df['HomeID'] = master_df['Home'].map(team_to_id).fillna(0).astype(int) # 9. JSON preparation: [Year, DayOfYear, VisitorID, HomeID] js_records = master_df[['Year', 'DayOfYear', 'VisitorID', 'HomeID']].values.tolist() # Era definitions for the sidebar era_definitions = [ ("All Time", range(1800, 2100)), ("Pre-modern Era (1876–1900)", range(1876, 1901)), ("Birth of Modern Era (1901–1919)", range(1901, 1920)), ("Segregation Era (1920–1946)", range(1920, 1947)), ("Integration Era (1947–1960)", range(1947, 1961)), ("First Expansion (1961–1968)", range(1961, 1969)), ("Birth of Division Play (1969–1993)", range(1969, 1994)), ("Wild Card Era (1994–Present)", range(1994, 2100)), ("30 Teams, 162 Games (1998–Present)", range(1998, 2100)) ] era_labels = [e[0] for e in era_definitions] era_ranges = [[min(e[1]), max(e[1])] for e in era_definitions] # Month/Day tooltips month_days = [] for i in range(1, 366): date_obj = datetime.strptime(f"2023 {i}", "%Y %j") month_days.append(date_obj.strftime("%m%d")) # JSON payloads - REVERSED years for the UI list records_json = json.dumps(js_records) years_json = json.dumps([int(y) for y in sorted(master_df['Year'].unique(), reverse=True)]) eras_json = json.dumps(era_labels) era_ranges_json = json.dumps(era_ranges) all_teams_json = json.dumps(all_teams_global) modern_teams_json = json.dumps(modern_teams if modern_teams else all_teams_global) month_days_json = json.dumps(month_days) # 10. GENERATE HTML final_html = f""" Baseball Schedule Analytics
CHART A
All Time
Total Games: 0
""" with open('game_frequency_chart.html', 'w', encoding='utf-8') as f: f.write(final_html) print(f"Data extraction refined: Official games count verified with improved cancellation filtering.") if __name__ == "__main__": analyze_game_schedules()