Refactor the parsing of schedule to get more info

src/load_options.py

@@ -11,7 +11,7 @@ from domain.playoffs import (
 )
 from domain.teams import SCHEDULE_NAME_TO_PFR_NAME_MAP
 from queries.nflverse.github_data import get_weekly_rosters
-from queries.pfr.league_schedule import get_season_time_map
+from queries.pfr.league_schedule import get_season_game_map
 
 
 @dataclass
@@ -99,8 +99,7 @@ def load_options():
     df_rosters = get_weekly_rosters()
 
     # get game schedules
-    week_game_times = get_season_time_map(SEASON)
-    latest_game_time_defaults = {k: max(v.values()) for k, v in week_game_times.items() if v}
+    week_game_times, latest_game_time_defaults = get_season_game_map(SEASON)
 
     # sort
     sort_by_cols = ["position", "fantasy_points", "week"]
@@ -117,9 +116,9 @@ def load_options():
     if len(df_rosters) == 0:
         return initialize_empty_options_map()
     df_rosters["gametime"] = df_rosters.apply(
-        lambda x: week_game_times.get(x.week, {}).get(
-            SCHEDULE_NAME_TO_PFR_NAME_MAP[x.team], latest_game_time_defaults.get(x.week, None)
-        ),
+        lambda x: week_game_times.get(x.week, {})
+        .get(SCHEDULE_NAME_TO_PFR_NAME_MAP[x.team], {})
+        .get("gametime", latest_game_time_defaults.get(x.week, None)),
         axis=1,
     )
 

src/queries/pfr/league_schedule.py

@@ -12,16 +12,35 @@ def get_full_schedule(season_int: str | int) -> pd.DataFrame:
     return df
 
 
-def get_week_team_time_map(df_schedule: pd.DataFrame) -> dict[int, dict[str, pd.Timestamp]]:
-    week_team_time_map: dict[int, dict[str, pd.Timestamp]] = {k: {} for k in SCHEDULE_WEEK_TO_PLAYOFF_WEEK.values()}
+def get_season_game_map(
+    season_int: str | int,
+) -> tuple[dict[int, dict[str, str | int | dict[str, pd.Timestamp]]], dict[int, pd.Timestamp]]:
+    df_schedule = get_full_schedule(season_int)
+    week_team_time_map: dict[int, dict[str, str | int | dict[str, pd.Timestamp]]] = {
+        k: {} for k in SCHEDULE_WEEK_TO_PLAYOFF_WEEK.values()
+    }
+    last_game_week_map: dict[int, pd.Timestamp] = {}
+
     for _, row in df_schedule.iterrows():
         if mapped_week := SCHEDULE_WEEK_TO_PLAYOFF_WEEK.get(row.Week):
             game_time = pd.to_datetime(row.Date + " " + row.Time, yearfirst=True).tz_localize("EST")
-            week_team_time_map[mapped_week][row["Winner/tie"]] = game_time
-            week_team_time_map[mapped_week][row["Loser/tie"]] = game_time
-    return week_team_time_map
 
-
-def get_season_time_map(season_int: str | int) -> dict[int, dict[str, pd.Timestamp]]:
-    df_schedule = get_full_schedule(season_int)
-    return get_week_team_time_map(df_schedule)
+            if mapped_week not in last_game_week_map:
+                last_game_week_map[mapped_week] = game_time
+            else:
+                last_game_week_map[mapped_week] = max(game_time, last_game_week_map[mapped_week])
+
+            # only actual winner/loser if game has already happened
+            winner_team = row["Winner/tie"]
+            loser_team = row["Loser/tie"]
+            winner_game = {
+                "gametime": game_time,
+                "opponent": loser_team,
+            }
+            loser_game = {
+                "gametime": game_time,
+                "opponent": winner_team,
+            }
+            week_team_time_map[mapped_week][winner_team] = winner_game
+            week_team_time_map[mapped_week][loser_team] = loser_game
+    return week_team_time_map, last_game_week_map