paul/baseball-nn
1
0
Fork 0
Code Issues 3 Pull Requests 1 Actions Packages Projects Releases Wiki Activity

Updating file parsing

Browse Source
This commit is contained in:
paul
2025-08-08 19:17:44 +00:00
parent bdaa432ab7
commit f59ae02734
2 changed files with 64 additions and 101 deletions
Download Patch File Download Diff File Expand all files Collapse all files

109
data/stats_importer.py
View File

@@ -29,13 +29,12 @@ class Importer:
src_file = os.path.join(source_dir, filename)
dest_file = os.path.join(dest_dir, filename)
self.parse_one_file(f"{source_dir}/{filename}")
try:
shutil.copy(src_file, dest_file)
print(f"Copied {filename} to {dest_dir}")
except Exception as e:
print(f"Failed to copy {filename}: {e}")
if self.parse_one_file(f"{source_dir}/{filename}"):
try:
shutil.copy(src_file, dest_file)
print(f"Copied {filename} to {dest_dir}")
except Exception as e:
print(f"Failed to copy {filename}: {e}")
def parse_one_file(self, filepath):
bb_dict = {}
@@ -44,15 +43,18 @@ class Importer:
bb_dict = list(reader)
for game in bb_dict:
self.populate_database_with_stats(game)
if not self.populate_database_with_stats(game):
return False
return True
def populate_database_with_stats(self, game_stats):
def populate_database_with_stats(self, game_stats) -> bool:
parkid = game_stats["park-id"]
park_data = self.database.select("SELECT latitude, longitude FROM parks WHERE park_id = ?", (game_stats["park-id"],))
if park_data is None:
print(f"{parkid} is None")
return
return True
insert_game = """
INSERT INTO games
@@ -79,19 +81,6 @@ class Importer:
)
"""
game_data = [
game_stats["date"], game_stats["num-of-game"], game_stats["day-of-week"],
game_stats["length-in-outs"], game_stats["day-night"], game_stats["completion-info"],
game_stats["forfeit"], game_stats["protest"], game_stats["park-id"],
game_stats["attendance"], game_stats["length-in-min"], game_stats["home-plate-ump-id"],
game_stats["home-plate-ump-name"], game_stats["1b-plate-ump-id"], game_stats["1b-plate-ump-name"],
game_stats["2b-plate-ump-id"], game_stats["2b-plate-ump-name"], game_stats["3b-plate-ump-id"],
game_stats["3b-plate-ump-name"], game_stats["lf-plate-ump-id"], game_stats["lf-plate-ump-name"],
game_stats["rf-plate-ump-id"], game_stats["rf-plate-ump-name"],
]
game_id = self.database.insert(insert_game, game_data)
insert_team_game = """
INSERT INTO team_game
(
@@ -145,6 +134,53 @@ class Importer:
)
"""
insert_into_weather = """
INSERT INTO weather
(
game_id, temperature, humidity,
dew_point, apparent_temperature, air_pressure,
precipitation, rain, snowfall,
cloud_cover, wind_speed, wind_direction,
wind_gusts, sun_rise, sun_set,
moon_phase
)
VALUES
(
?, ?, ?,
?, ?, ?,
?, ?, ?,
?, ?, ?,
?, ?, ?,
?
)
"""
hour = 15 if game_stats["day-night"] == "D" else 19
historic_weather = get_weather(park_data[0], park_data[1], game_stats["date"], hour)
if "error" in historic_weather:
print(f"Error: {historic_weather['error']}: Details: {historic_weather['details']}")
return False
if "hourly" not in historic_weather:
print(f"Failed to get weather: Full JSON: {historic_weather}")
return False
historic_weather = historic_weather["hourly"]
game_data = [
game_stats["date"], game_stats["num-of-game"], game_stats["day-of-week"],
game_stats["length-in-outs"], game_stats["day-night"], game_stats["completion-info"],
game_stats["forfeit"], game_stats["protest"], game_stats["park-id"],
game_stats["attendance"], game_stats["length-in-min"], game_stats["home-plate-ump-id"],
game_stats["home-plate-ump-name"], game_stats["1b-plate-ump-id"], game_stats["1b-plate-ump-name"],
game_stats["2b-plate-ump-id"], game_stats["2b-plate-ump-name"], game_stats["3b-plate-ump-id"],
game_stats["3b-plate-ump-name"], game_stats["lf-plate-ump-id"], game_stats["lf-plate-ump-name"],
game_stats["rf-plate-ump-id"], game_stats["rf-plate-ump-name"],
]
game_id = self.database.insert(insert_game, game_data)
visiting_team_data = [
game_id, game_stats["visiting-team"], game_stats["visiting-game-num"],
game_stats["visiting-score"], game_stats["visiting-line-scores"], game_stats["visiting-at-bats"],
@@ -198,31 +234,6 @@ class Importer:
self.database.insert(insert_team_game, visiting_team_data)
self.database.insert(insert_team_game, home_team_data)
hour = 15 if game_stats["day-night"] == "D" else 19
historic_weather = get_weather(park_data[0], park_data[1], game_stats["date"], hour)
historic_weather = historic_weather["hourly"]
insert_into_weather = """
INSERT INTO weather
(
game_id, temperature, humidity,
dew_point, apparent_temperature, air_pressure,
precipitation, rain, snowfall,
cloud_cover, wind_speed, wind_direction,
wind_gusts, sun_rise, sun_set,
moon_phase
)
VALUES
(
?, ?, ?,
?, ?, ?,
?, ?, ?,
?, ?, ?,
?, ?, ?,
?
)
"""
(sunrise_time, sunset_time, moonphase) = get_sun_and_moon_phase(park_data[0], park_data[1], game_stats["date"])
weather_data = [
@@ -235,3 +246,5 @@ class Importer:
]
self.database.insert(insert_into_weather, weather_data)
return True