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

Adding win and if the team is home to the database, calculations for win and lose streak, and returning the training data. Fixing sunrise sunset times

Browse Source
This commit is contained in:
paul
2025-08-29 20:09:49 +00:00
parent 330c7445a6
commit 3950e99151
7 changed files with 221 additions and 144 deletions
Download Patch File Download Diff File Expand all files Collapse all files

21
data/build_weather.py
View File

@@ -1,21 +1,20 @@
import ephem
import requests
import datetime
import requests
from suntime import Sun, SunTimeException
from datetime import datetime
from dateutil import tz
from timezonefinder import TimezoneFinder
def get_sun_and_moon_phase(lat, long, date_str):
curr_date = datetime.strptime(date_str, "%Y%m%d")
date_str = f"{date_str[:4]}/{date_str[4:6]}/{date_str[6:8]}"
observer = ephem.Observer()
observer.lat = str(lat)
observer.lon = str(long)
observer.date = date_str
sun = ephem.Sun()
sun.compute(observer)
sunrise_time = int(observer.next_rising(sun).datetime().strftime('%Y%m%d'))
sunset_time = int(observer.next_setting(sun).datetime().strftime('%Y%m%d'))
my_tz = get_timezone(lat, long)
sun = Sun(lat, long)
sunrise_time = sun.get_sunrise_time(curr_date, tz.gettz(my_tz)).strftime("%H:%M:%S")
sunset_time = sun.get_sunset_time(curr_date, tz.gettz(my_tz)).strftime("%H:%M:%S")
date = ephem.Date(date_str)
moon = ephem.Moon()

189
data/get_data.py
View File

@@ -211,7 +211,7 @@ def pull_training_data(database: Database, game_date: str, game_number: int, par
- starting_9_position
- Output
- 0 home team wins, 1 visiting team wins
- 1 home team wins, 0 visiting team wins
"""
select_upcoming_game = """
@@ -257,6 +257,7 @@ def pull_training_data(database: Database, game_date: str, game_number: int, par
select_teams = """
SELECT
win, team, home,
game_num, manager_id, starting_pitcher_id,
starting_1_id, starting_1_position,
starting_2_id, starting_2_position,
@@ -267,76 +268,138 @@ def pull_training_data(database: Database, game_date: str, game_number: int, par
starting_7_id, starting_7_position,
starting_8_id, starting_8_position,
starting_9_id, starting_9_position
FROM
team_game
WHERE
game = ?
"""
curr_team_game = database.selectall(select_teams, [curr_game[0]])
if curr_game.len() != 2:
print(f"Got the wrong number of games {curr_game.len()}")
if len(curr_team_game) != 2:
print(f"Got the wrong number of games {len(curr_team_game)}")
return None
training_result = (curr_team_game[0][0] == 1 and curr_team_game[0][2] == 1)
select_previous_games = """
SELECT
- win_streak
- loss_streak
- game_date
- day_of_weeki
- length_in_outs
- day_night
- completion_info
- forfeit
- protest
- park_id
- attendence
- length_in_minutes
- hits
- doubles
- triples
- homeruns
- rbis
- sacrifice_hits
- sacrifice_flies
- hit_by_pitch
- walks
- intentional_walks
- strikeouts
- stolen_bases
- caught_stealing
- grounded_double
- interference
- left_on_base
- pitchers_used
- individual_earned_runs
- earned_runs
- wild_pitches
- balks
- putouts
- assists
- errors
- passed
- double_play
- triple_play
- game_num
- manager_id
- starting_pitcher_id
- starting_1_id
- starting_1_position
- starting_2_id
- starting_2_position
- starting_3_id
- starting_3_position
- starting_4_id
- starting_4_position
- starting_5_id
- starting_5_position
- starting_6_id
- starting_6_position
- starting_7_id
- starting_7_position
- starting_8_id
- starting_8_position
- starting_9_id
- starting_9_position
games.id,
games.game_date, games.game_number,
games.day_of_week, games.day_night,
games.length_in_outs, games.completion_info,
games.forfeit, games.protest,
games.attendence, games.length_in_minutes,
games.home_plate_ump_id,
games.b1_ump_id, games.b2_ump_id, games.b3_ump_id,
games.lf_ump_id, games.rf_ump_id,
team_game.game_num, team_game.score,
team_game.line_score, team_game.win,
team_game.hits, team_game.doubles, team_game.triples,
team_game.homeruns, team_game.rbis,
team_game.sacrifice_hits, team_game.sacrifice_flies,
team_game.hit_by_pitch, team_game.walks, team_game.intentional_walks,
team_game.strikeouts, team_game.stolen_bases,
team_game.caught_stealing, team_game.grounded_double,
team_game.interference, team_game.left_on_base,
team_game.pitchers_used,
team_game.individual_earned_runs, team_game.earned_runs,
team_game.wild_pitches, team_game.balks,
team_game.putouts, team_game.assists,
team_game.errors, team_game.passed,
team_game.double_play, team_game.triple_play,
team_game.manager_id, team_game.starting_pitcher_id,
team_game.starting_1_id, team_game.starting_1_position,
team_game.starting_2_id, team_game.starting_2_position,
team_game.starting_3_id, team_game.starting_3_position,
team_game.starting_4_id, team_game.starting_4_position,
team_game.starting_5_id, team_game.starting_5_position,
team_game.starting_6_id, team_game.starting_6_position,
team_game.starting_7_id, team_game.starting_7_position,
team_game.starting_8_id, team_game.starting_8_position,
team_game.starting_9_id, team_game.starting_9_position,
parks.park_id,
weather.temperature, weather.humidity, weather.dew_point,
weather.apparent_temperature, weather.air_pressure,
weather.precipitation, weather.rain, weather.snowfall,
weather.cloud_cover,
weather.wind_speed, weather.wind_direction, weather.wind_gusts,
weather.sun_rise, weather.sun_set, weather.moon_phase
FROM
games
LEFT JOIN parks ON parks.park_id = games.park_id
LEFT JOIN team_game ON team_game.game = games.id
LEFT JOIN weather ON weather.game_id = games.id
WHERE
games.game_date > ? AND
games.game_date < ? AND
team_game.team = ?
ORDER BY team_game.game_num ASC
"""
first_of_the_year = f"{game_date[:4]}0101"
select_previous_games_data_0 = [first_of_the_year, game_date, curr_team_game[0][1]]
select_previous_games_data_1 = [first_of_the_year, game_date, curr_team_game[1][1]]
training_data = [*curr_game[1:]]
training_data = [*training_data, *curr_team_game[0][1:]]
prev_games = database.selectall(select_previous_games, select_previous_games_data_0)
prev_game_data = [0] * 61
if prev_games is not None:
prev_win_streak = 0
index = len(prev_games) - 1
while index > 0 and prev_games[index][20] == 1:
prev_win_streak += 1
index -= 1
prev_loss_streak = 0
index = len(prev_games) - 1
while index > 0 and prev_games[index][20] == 0:
prev_loss_streak += 1
index -= 1
index = len(prev_games) - 1
prev_game_data = [
prev_win_streak,
prev_loss_streak,
*prev_games[index][1:]
]
training_data = [*training_data, *prev_game_data]
training_data = [*training_data, *curr_team_game[1][1:]]
prev_games = database.selectall(select_previous_games, select_previous_games_data_1)
prev_game_data = [0] * 61
if prev_games is not None:
prev_win_streak = 0
index = len(prev_games) - 1
while index > 0 and prev_games[index][20] == 1:
prev_win_streak += 1
index -= 1
prev_loss_streak = 0
index = len(prev_games) - 1
while index > 0 and prev_games[index][20] == 0:
prev_loss_streak += 1
index -= 1
index = len(prev_games) - 1
prev_game_data = [
prev_win_streak,
prev_loss_streak,
*prev_games[index][1:]
]
training_data = [*training_data, *prev_game_data]
print(f"{training_result}\n{training_data}")
return (training_result, training_data)

4
data/sql/build_db.sql
View File

@@ -2,7 +2,7 @@ CREATE TABLE IF NOT EXISTS parks (
park_id CHAR(5) NOT NULL,
park_name VARCHAR,
lf_fence_distance SMALLINT UNSIGNED,
lf_fence_height, SMALLINT UNSIGNED,
lf_fence_height SMALLINT UNSIGNED,
ct_fence_distance SMALLINT UNSIGNED,
ct_fence_height SMALLINT UNSIGNED,
rf_fence_distance SMALLINT UNSIGNED,
@@ -59,6 +59,8 @@ CREATE TABLE IF NOT EXISTS team_game (
game_num TINYINT UNSIGNED,
score TINYINT UNSIGNED,
line_score CHAR(9),
win BIT,
home BIT,
at_bats TINYINT UNSIGNED,
hits TINYINT UNSIGNED,

10
data/stats_importer.py
View File

@@ -107,7 +107,7 @@ class Importer:
INSERT INTO team_game
(
game, team, game_num,
score, line_score, at_bats,
score, line_score, win, home, at_bats,
hits, doubles, triples,
homeruns, rbis, sacrifice_hits,
sacrifice_flies, hit_by_pitch, walks,
@@ -132,7 +132,7 @@ class Importer:
VALUES
(
?, ?, ?,
?, ?, ?,
?, ?, ?, ?, ?,
?, ?, ?,
?, ?, ?,
?, ?, ?,
@@ -205,9 +205,10 @@ class Importer:
game_id = self.database.insert(insert_game, game_data)
visiting_win_loss = int(game_stats["visiting-score"]) > int(game_stats["home-score"])
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"],
game_stats["visiting-score"], game_stats["visiting-line-scores"], visiting_win_loss, 0, game_stats["visiting-at-bats"],
game_stats["visiting-hits"], game_stats["visiting-doubles"], game_stats["visiting-triples"],
game_stats["visiting-homeruns"], game_stats["visiting-rbi"], game_stats["visiting-sacrifice-hits"],
game_stats["visiting-sacrifice-flies"], game_stats["visiting-hit-by-pitch"], game_stats["visiting-walks"],
@@ -230,9 +231,10 @@ class Importer:
game_stats["visiting-starting-9-position"]
]
home_win_loss = int(game_stats["home-score"]) > int(game_stats["visiting-score"])
home_team_data = [
game_id, game_stats["home-team"], game_stats["home-game-num"],
game_stats["home-score"], game_stats["home-line-scores"], game_stats["home-at-bats"],
game_stats["home-score"], game_stats["home-line-scores"], home_win_loss, 1, game_stats["home-at-bats"],
game_stats["home-hits"], game_stats["home-doubles"], game_stats["home-triples"],
game_stats["home-homeruns"], game_stats["home-rbi"], game_stats["home-sacrifice-hits"],
game_stats["home-sacrifice-flies"], game_stats["home-hit-by-pitch"], game_stats["home-walks"],