def parse_boxscores(predictor, teams, skip_save_to_mongodb):
games_list = []
match_info = []
prediction_stats = []
stdev_dict = find_stdev_for_every_stat(teams)
today = datetime.today()
today_string = '%s-%s-%s' % (today.month, today.day, today.year)
for game in Boxscores(today).games[today_string]:
# Skip the games that are not between two DI teams since stats are not
# saved for those teams.
if game['non_di']:
continue
for sim in range(NUM_SIMS):
home = Team(game['home_name'], game['home_abbr'])
away = Team(game['away_name'], game['away_abbr'])
title = '%s at %s' % (away.name, home.name)
game_info = GameInfo(home, away, title)
games_list.append(game_info)
match_stats = get_match_stats(game, stdev_dict)
prediction_stats.append(match_stats)
home_name = game['home_name']
away_name = game['away_name']
if game['home_rank']:
home_name = '(%s) %s' % (game['home_rank'], home_name)
if game['away_rank']:
away_name = '(%s) %s' % (game['away_rank'], away_name)
g = MatchInfo(away_name, home_name, game['away_abbr'],
game['home_abbr'], game['top_25'], None, match_stats)
match_info.append(g)
predictions = make_predictions(prediction_stats, games_list, match_info,
predictor)
save_predictions(predictions, skip_save_to_mongodb)
def get_scores(year, month ,day):
home_teams = []
away_teams = []
home_scores = []
away_scores = []
games = Boxscores(datetime(year, month, day))
all_games = games.games[str(month)+'-'+str(day)+'-'+str(year)]
for game in all_games:
home_teams.append(game['home_name'])
away_teams.append(game['away_name'])
home_scores.append(game['home_score'])
away_scores.append(game['away_score'])
name = str(year)+"_"+str(month)+"_"+str(day)+"_scores.csv"
with open(name, 'w', newline='') as file:
writer = csv.writer(file)
writer.writerow(["home_team","away_team","home_score","away_score"])
for i,team in enumerate(home_teams):
writer.writerow([home_teams[i],away_teams[i],home_scores[i],away_scores[i]])
path = os.getcwd()
os.rename(path+"\\"+name, path+"\\"+"scores\\"+name)
return home_teams,away_teams,home_scores,away_scores
def get_yesterdays_scores():
today = datetime.today() - timedelta(days=1)
date = datetime.today().strftime('%Y%m%d')
games_today = Boxscores(today).games[today.strftime('%-m-%-d-%Y')]
output = []
for g in games_today:
if g['away_score'] == 'None':
print('Not all scores available yet')
break
else:
try:
output.append({
'date': date,
'away': team_alt_names[g['away_name']],
'home': team_alt_names[g['home_name']],
"away_score": g['away_score'],
"home_score": g['home_score']
})
except:
print('Cant find team name in matchup: {} v. {}'.format(
g['away_name'], g['home_name']))
outcome = pd.DataFrame(data=output)
outcome["key"] = outcome["date"] + outcome["home"]
outcome = outcome[[
'key', 'date', 'away', 'home', "away_score", "home_score"
]]
td.append_df_to_excel('Historic_Scores', outcome)
def find_todays_games(teams):
games_list = []
today = datetime.today()
today_string = '%s-%s-%s' % (today.month, today.day, today.year)
for game in Boxscores(today).games[today_string]:
# Skip the games that are not between two DI teams since stats are not
# saved for those teams.
if game['non_di']:
continue
games_list.append(populate_game_info(teams, game))
return games_list
def iterate_files(files):
total_games = 0
total_correct = 0
for filename in files:
date = get_date(filename)
saved_data = get_saved_prediction(filename)
games = Boxscores(date).games[filename.replace('.json', '')]
saved_data, num_games, num_correct = parse_boxscore(games, saved_data)
total_games += num_games
total_correct += num_correct
save_json(saved_data, 'predictions/%s' % filename)
print('=' * 80)
print(' Accuracy: %s%%' % round(100.0 * float(total_correct) / \
float(total_games), 2))
print('=' * 80)
class TestNCAABBoxscores:
@patch('requests.get', side_effect=mock_pyquery)
def setup_method(self, *args, **kwargs):
flexmock(Boxscores) \
.should_receive('_find_games') \
.and_return(None)
self.boxscores = Boxscores(None)
def test_boxscore_with_no_score_returns_none(self):
mock_html = pq("""<table class="teams">
<tbody>
<tr class="loser">
<td><a href="/cbb/schools/south-dakota/2019.html">South Dakota</a></td>
<td class="right"></td>
<td class="right gamelink">
</td>
</tr>
<tr class="loser">
<td><a href="/cbb/schools/kansas/2019.html">Kansas</a>\
<span class='pollrank'> (1) </span></td>
<td class="right"></td>
<td class="right">
</td>
</tr>
</tbody>
</table>""")
games = self.boxscores._extract_game_info([mock_html])
assert games == [
{
'home_name': 'Kansas',
'home_abbr': 'kansas',
'away_name': 'South Dakota',
'away_abbr': 'south-dakota',
'boxscore': '',
'non_di': False,
'top_25': True,
'home_score': None,
'home_rank': 1,
'away_score': None,
'away_rank': None,
'winning_name': None,
'winning_abbr': None,
'losing_name': None,
'losing_abbr': None
}
]
def main():
min_date = date(2021, 11, 9)
max_date = date(2022, 3, 14)
print("Fetching Games")
boxs = Boxscores(min_date, max_date)
print("Writing Games")
with open('../../data/gamelist_20220314.pkl', 'wb') as fh:
pickle.dump(boxs.games, fh)
print("Fetching Teams")
teams = Teams()
print("Writing Teams")
with open('../../data/teamlist_20220314.pkl', 'wb') as fh:
pickle.dump(teams, fh)
def test_boxscores_search(self, *args, **kwargs):
expected = {
'boxscores': [
{
'home_name': 'Buffalo',
'away_abbr': 'canisius',
'home_abbr': 'buffalo',
'boxscore': '2017-11-11-13-buffalo',
'away_name': 'Canisius',
'non_di': False
},
{
'home_name': 'Fairfield',
'away_abbr': 'pennsylvania',
'home_abbr': 'fairfield',
'boxscore': '2017-11-11-13-fairfield',
'away_name': 'Penn',
'non_di': False
},
{
'home_name': 'Florida Gulf Coast',
'away_abbr': 'illinois-state',
'home_abbr': 'florida-gulf-coast',
'boxscore': '2017-11-11-13-florida-gulf-coast',
'away_name': 'Illinois State',
'non_di': False
},
{
'home_name': 'Bradley',
'away_abbr': 'iupui',
'home_abbr': 'bradley',
'boxscore': '2017-11-11-14-bradley',
'away_name': 'IUPUI',
'non_di': False
},
{
'home_name': 'Ohio',
'away_abbr': 'alabama-am',
'home_abbr': 'ohio',
'boxscore': '2017-11-11-14-ohio',
'away_name': 'Alabama A&M',
'non_di': False
},
{
'home_name': 'Quinnipiac',
'away_abbr': 'dartmouth',
'home_abbr': 'quinnipiac',
'boxscore': '2017-11-11-14-quinnipiac',
'away_name': 'Dartmouth',
'non_di': False
},
{
'home_name': 'Western Michigan',
'away_abbr': 'Siena Heights',
'home_abbr': 'western-michigan',
'boxscore': '2017-11-11-14-western-michigan',
'away_name': 'Siena Heights',
'non_di': True
},
{
'home_name': 'Appalachian State',
'away_abbr': 'Toccoa Falls',
'home_abbr': 'appalachian-state',
'boxscore': '2017-11-11-15-appalachian-state',
'away_name': 'Toccoa Falls',
'non_di': True
},
{
'home_name': 'Brown',
'away_abbr': 'Johnson & Wales (RI)',
'home_abbr': 'brown',
'boxscore': '2017-11-11-15-brown',
'away_name': 'Johnson & Wales (RI)',
'non_di': True
},
{
'home_name': 'Drake',
'away_abbr': 'Coe',
'home_abbr': 'drake',
'boxscore': '2017-11-11-15-drake',
'away_name': 'Coe',
'non_di': True
},
{
'home_name': 'La Salle',
'away_abbr': 'saint-peters',
'home_abbr': 'la-salle',
'boxscore': '2017-11-11-15-la-salle',
'away_name': 'St. Peter\'s',
'non_di': False
},
{
'home_name': 'Longwood',
'away_abbr': 'stephen-f-austin',
'home_abbr': 'longwood',
'boxscore': '2017-11-11-15-longwood',
'away_name': 'Stephen F. Austin',
'non_di': False
},
{
'home_name': 'Little Rock',
'away_abbr': 'Ouachita',
'home_abbr': 'arkansas-little-rock',
'boxscore': '2017-11-11-16-arkansas-little-rock',
'away_name': 'Ouachita',
'non_di': True
},
{
'home_name': 'UCSB',
'away_abbr': 'north-dakota-state',
'home_abbr': 'california-santa-barbara',
'boxscore': '2017-11-11-16-california-santa-barbara',
'away_name': 'North Dakota State',
'non_di': False
},
{
'home_name': 'DePaul',
'away_abbr': 'notre-dame',
'home_abbr': 'depaul',
'boxscore': '2017-11-11-16-depaul',
'away_name': 'Notre Dame',
'non_di': False
},
{
'home_name': 'Duquesne',
'away_abbr': 'st-francis-ny',
'home_abbr': 'duquesne',
'boxscore': '2017-11-11-17-duquesne',
'away_name': 'St. Francis (NY)',
'non_di': False
},
{
'home_name': 'NJIT',
'away_abbr': 'wagner',
'home_abbr': 'njit',
'boxscore': '2017-11-11-18-njit',
'away_name': 'Wagner',
'non_di': False
},
{
'home_name': 'Akron',
'away_abbr': 'cleveland-state',
'home_abbr': 'akron',
'boxscore': '2017-11-11-19-akron',
'away_name': 'Cleveland State',
'non_di': False
},
{
'home_name': 'Duke',
'away_abbr': 'utah-valley',
'home_abbr': 'duke',
'boxscore': '2017-11-11-19-duke',
'away_name': 'Utah Valley',
'non_di': False
},
{
'home_name': 'Elon',
'away_abbr': 'Peace',
'home_abbr': 'elon',
'boxscore': '2017-11-11-19-elon',
'away_name': 'Peace',
'non_di': True
},
{
'home_name': 'Marist',
'away_abbr': 'lehigh',
'home_abbr': 'marist',
'boxscore': '2017-11-11-19-marist',
'away_name': 'Lehigh',
'non_di': False
},
{
'home_name': 'Michigan',
'away_abbr': 'north-florida',
'home_abbr': 'michigan',
'boxscore': '2017-11-11-19-michigan',
'away_name': 'North Florida',
'non_di': False
},
{
'home_name': 'North Carolina-Wilmington',
'away_abbr': 'North Carolina Wesleyan',
'home_abbr': 'north-carolina-wilmington',
'boxscore': '2017-11-11-19-north-carolina-wilmington',
'away_name': 'North Carolina Wesleyan',
'non_di': True
},
{
'home_name': 'Presbyterian',
'away_abbr': 'Johnson University',
'home_abbr': 'presbyterian',
'boxscore': '2017-11-11-19-presbyterian',
'away_name': 'Johnson University',
'non_di': True
},
{
'home_name': 'Sam Houston State',
'away_abbr': 'Texas-Tyler',
'home_abbr': 'sam-houston-state',
'boxscore': '2017-11-11-19-sam-houston-state',
'away_name': 'Texas-Tyler',
'non_di': True
},
{
'home_name': 'South Dakota',
'away_abbr': 'Mayville State',
'home_abbr': 'south-dakota',
'boxscore': '2017-11-11-19-south-dakota',
'away_name': 'Mayville State',
'non_di': True
},
{
'home_name': 'Toledo',
'away_abbr': 'saint-josephs',
'home_abbr': 'toledo',
'boxscore': '2017-11-11-19-toledo',
'away_name': 'St. Joseph\'s',
'non_di': False
},
{
'home_name': 'Nebraska',
'away_abbr': 'eastern-illinois',
'home_abbr': 'nebraska',
'boxscore': '2017-11-11-20-nebraska',
'away_name': 'Eastern Illinois',
'non_di': False
},
{
'home_name': 'Saint Mary\'s (CA)',
'away_abbr': 'saint-francis-pa',
'home_abbr': 'saint-marys-ca',
'boxscore': '2017-11-11-20-saint-marys-ca',
'away_name': 'Saint Francis (PA)',
'non_di': False
},
{
'home_name': 'Texas-Arlington',
'away_abbr': 'loyola-marymount',
'home_abbr': 'texas-arlington',
'boxscore': '2017-11-11-20-texas-arlington',
'away_name': 'Loyola Marymount',
'non_di': False
},
{
'home_name': 'Western Illinois',
'away_abbr': 'Saint Mary\'s (MN)',
'home_abbr': 'western-illinois',
'boxscore': '2017-11-11-20-western-illinois',
'away_name': 'Saint Mary\'s (MN)',
'non_di': True
},
{
'home_name': 'BYU',
'away_abbr': 'mississippi-valley-state',
'home_abbr': 'brigham-young',
'boxscore': '2017-11-11-21-brigham-young',
'away_name': 'Mississippi Valley State',
'non_di': False
},
{
'home_name': 'Kent State',
'away_abbr': 'youngstown-state',
'home_abbr': 'kent-state',
'boxscore': '2017-11-11-21-kent-state',
'away_name': 'Youngstown State',
'non_di': False
},
{
'home_name': 'New Mexico',
'away_abbr': 'Northern New Mexico',
'home_abbr': 'new-mexico',
'boxscore': '2017-11-11-21-new-mexico',
'away_name': 'Northern New Mexico',
'non_di': True
},
{
'home_name': 'UNLV',
'away_abbr': 'florida-am',
'home_abbr': 'nevada-las-vegas',
'boxscore': '2017-11-11-22-nevada-las-vegas',
'away_name': 'Florida A&M',
'non_di': False
},
{
'home_name': 'Portland',
'away_abbr': 'portland-state',
'home_abbr': 'portland',
'boxscore': '2017-11-11-22-portland',
'away_name': 'Portland State',
'non_di': False
},
]
}
result = Boxscores(datetime(2017, 11, 11)).games
assert result == expected
def projections():
home = request.args.get('home')
away = request.args.get('away')
if home is None and away is None:
today = dt.today()
today_start = today.replace(hour=0, minute=0, second=0, microsecond=0)
today_end = today_start + timedelta(1)
games = db.session.query(Game).filter(Game.date > today_start,
Game.date < today_end).all()
all_games = db.session.query(Game).all()
if len(all_games) > 9500:
db.session.query(Game).delete()
db.session.commit()
# If today's games have not been retrieved yet then do so from sports
# And get the projected score from the data science API
if len(games) == 0:
games_today = Boxscores(dt.today())
games_today = games_today.games
rankings = Rankings()
rankings = rankings.current
for date in games_today.keys():
for game in games_today[date]:
home_team = db.session.query(Team).filter(
Team.name == game["home_name"]).first()
away_team = db.session.query(Team).filter(
Team.name == game["away_name"]).first()
if home_team is None:
home_team = Team(id=game["home_abbr"],
name=game["home_name"])
db.session.add(home_team)
try:
db.session.commit()
home_team = db.session.query(Team).filter(
Team.name == game["home_name"]).first()
except:
db.session.rollback()
elif away_team is None:
away_team = Team(id=game["away_abbr"],
name=game["away_name"])
db.session.add(away_team)
try:
db.session.commit()
away_team = db.session.query(Team).filter(
Team.name == game["away_name"]).first()
except:
db.session.rollback()
print("000ooof")
if home_team and away_team:
data = {
"year":
dt.today().year,
"month":
dt.today().month,
"day":
dt.today().day,
"home_name":
home_team.name,
"home_rank":
float(rankings.get(home_team.id.lower(), 0)),
"away_name":
away_team.name,
"away_rank":
float(rankings.get(away_team.id.lower(), 0))
}
res = requests.post("http://ncaab.herokuapp.com/",
data=json.dumps(data))
res = res.json()
home_projection = res["home_score"]
away_projection = res["away_score"]
new_game = Game(date=date,
home=home_team.id,
away=away_team.id,
home_projection=home_projection,
away_projection=away_projection)
db.session.add(new_game)
games.append(new_game)
# TODO: Test this out...
# Update yesterday's games with the scores
# yesterday_start = today_start - timedelta(1)
# games_yesterday = Boxscores(yesterday_start, today_start)
# games_yesterday = games_yesterday.games
# for date in games_yesterday.keys():
# for game in games_yesterday[date]:
# home_team = db.session.query(Team).filter(
# Team.name == game["home_name"]).first()
# away_team = db.session.query(Team).filter(
# Team.name == game["away_name"]).first()
# print(home_team, away_team)
# applicable_game = db.session.query(Game).filter(
# Game.date > yesterday_start, Game.date < today_start, Game.home == home_team.id, Game.away == away_team.id).first()
# print(applicable_game)
# applicable_game.home_score = game["home_score"]
# applicable_game.away_score = game["away_score"]
try:
db.session.commit()
except:
db.session.rollback()
games_without_names = [g.serialize() for g in games]
games_with_names = []
for game in games_without_names:
home_team = db.session.query(Team).filter(
Team.id == game["home_id"]).first()
away_team = db.session.query(Team).filter(
Team.id == game["away_id"]).first()
print("oof")
game["home_name"] = home_team.name
game["away_name"] = away_team.name
games_with_names.append(game)
return jsonify(games_with_names)
elif home is None or away is None:
return (
"'home' and 'away' are necessary query parameters when not using the 'all' query parameter"
)
else:
rankings = Rankings()
rankings = rankings.current
print(home)
print(away)
home_team = db.session.query(Team).filter(Team.id == home).first()
away_team = db.session.query(Team).filter(Team.id == away).first()
print("ooooof")
data = {
"year": dt.today().year,
"month": dt.today().month,
"day": dt.today().day,
"home_name": home_team.name,
"home_rank": float(rankings.get(home_team.id.lower(), 0)),
"away_name": away_team.name,
"away_rank": float(rankings.get(away_team.id.lower(), 0))
}
res = requests.post("http://ncaab.herokuapp.com/",
data=json.dumps(data))
res = res.json()
home_projection = res["home_score"]
away_projection = res["away_score"]
print("ummmmm")
return jsonify({
'home_id': home_team.id,
'away_id': away_team.id,
'home_name': home_team.name,
'away_name': away_team.name,
'home_projection': home_projection,
'away_projection': away_projection
})
datetime.date(2014, 3, 21),
datetime.date(2013, 3, 21),
datetime.date(2013, 3, 22),
datetime.date(2012, 3, 15),
datetime.date(2012, 3, 16),
datetime.date(2011, 3, 17),
datetime.date(2011, 3, 18),
datetime.date(2010, 3, 18),
datetime.date(2010, 3, 19),
datetime.date(2009, 3, 19),
datetime.date(2009, 3, 20)
]
sSum = np.zeros(16)
sCount = np.zeros(16)
for dDate in dates_firstRound:
gameList = Boxscores(dDate).games
for day in gameList:
for game in gameList[day]:
hRank = game['home_rank']
vRank = game['away_rank']
hTeam = game['home_name']
vTeam = game['away_name']
hScore = game['home_score']
vScore = game['away_score']
if not (hScore is None or vScore is None):
hDiff = hScore - vScore
vDiff = -hDiff
else:
continue
if not (hRank is None or vRank is None):
sSum[hRank - 1] += hDiff
def setup_method(self, *args, **kwargs):
flexmock(Boxscores) \
.should_receive('_find_games') \
.and_return(None)
self.boxscores = Boxscores(None)
def ncaab_boxscores(game_date: Date):
date_key = f"{game_date.month}-{game_date.day}-{game_date.year}"
return Boxscores(game_date).games[date_key]
def get_todays_games():
todaysGames = []
teamData = get_team_data()
today = datetime.today()
games_today = Boxscores(today).games[today.strftime('%-m-%-d-%Y')]
for game in games_today:
try:
awayName = team_alt_names[game['away_name']]
homeName = team_alt_names[game['home_name']]
awayInfo = teamData[awayName]
homeInfo = teamData[homeName]
ag = awayInfo.games_played
hg = homeInfo.games_played
away = ballgame(
awayName,
awayInfo.points / ag,
awayInfo.field_goals / ag,
awayInfo.field_goal_attempts / ag,
awayInfo.three_point_field_goals / ag,
awayInfo.three_point_field_goal_attempts / ag,
awayInfo.two_point_field_goals / ag,
awayInfo.two_point_field_goal_attempts / ag,
awayInfo.free_throws / ag,
awayInfo.free_throw_attempts / ag,
awayInfo.offensive_rebounds / ag,
awayInfo.defensive_rebounds / ag,
awayInfo.assists / ag,
awayInfo.turnovers / ag,
awayInfo.steals / ag,
awayInfo.field_goal_percentage,
awayInfo.free_throw_percentage,
homeInfo.opp_points / hg,
homeInfo.opp_field_goals / hg,
homeInfo.opp_field_goal_attempts / hg,
homeInfo.opp_three_point_field_goal_attempts / hg,
homeInfo.opp_three_point_field_goals / hg,
homeInfo.opp_two_point_field_goal_attempts / hg,
homeInfo.opp_free_throws / hg,
homeInfo.opp_free_throw_attempts / hg,
homeInfo.opp_defensive_rebounds / hg,
homeInfo.opp_assists / hg,
homeInfo.opp_turnovers / hg,
homeInfo.opp_steals / hg,
homeInfo.opp_blocks / hg,
homeInfo.opp_personal_fouls / hg,
homeInfo.opp_field_goal_percentage,
homeInfo.opp_three_point_field_goal_percentage,
homeInfo.opp_free_throw_percentage,
)
home = ballgame(
homeName,
homeInfo.points / hg,
homeInfo.field_goals / hg,
homeInfo.field_goal_attempts / hg,
homeInfo.three_point_field_goals / hg,
homeInfo.three_point_field_goal_attempts / hg,
homeInfo.two_point_field_goals / hg,
homeInfo.two_point_field_goal_attempts / hg,
homeInfo.free_throws / hg,
homeInfo.free_throw_attempts / hg,
homeInfo.offensive_rebounds / hg,
homeInfo.defensive_rebounds / hg,
homeInfo.assists / hg,
homeInfo.turnovers / hg,
homeInfo.steals / hg,
homeInfo.field_goal_percentage,
homeInfo.free_throw_percentage,
awayInfo.opp_points / ag,
awayInfo.opp_field_goals / ag,
awayInfo.opp_field_goal_attempts / ag,
awayInfo.opp_three_point_field_goal_attempts / ag,
awayInfo.opp_three_point_field_goals / ag,
awayInfo.opp_two_point_field_goal_attempts / ag,
awayInfo.opp_free_throws / ag,
awayInfo.opp_free_throw_attempts / ag,
awayInfo.opp_defensive_rebounds / ag,
awayInfo.opp_assists / ag,
awayInfo.opp_turnovers / ag,
awayInfo.opp_steals / ag,
awayInfo.opp_blocks / ag,
awayInfo.opp_personal_fouls / ag,
awayInfo.opp_field_goal_percentage,
awayInfo.opp_three_point_field_goal_percentage,
awayInfo.opp_free_throw_percentage,
)
todaysGames.append([away, home])
except:
print(game['away_name'], game['home_name'])
return todaysGames