def team_preprocessing():
"""
Get the Data and process it how I want it
:return DataFrame
"""
cols = [
"Team", "Season", "Game.ID", "Date", "Opponent", "Venue", "TOI", "GF",
'GA'
]
# Get Game Data for Teams
df = pd.read_csv("./data/teams_all_sits.csv")
for team_col in ['Team', "Opponent", "Venue"]:
df = helpers.fix_team(df, team_col)
df = df[cols]
# Get "correct" game id
df['game_id'] = df.apply(
lambda x: int(str(x['Season']) + "0" + str(x['Game.ID'])), axis=1)
df = df.drop(['Game.ID'], axis=1)
# Sort by game_id...lowest is first
df = df.sort_values(by=['game_id'])
# Merge in game outcomes:
df = helpers.merge_outcomes(df)
# Only keeps games from the home team perspective!!!!!!!!
df = df[df['Team'] == df['Venue']]
return df
def get_pp_data():
"""
Get and prepare the power play data
:return: DataFrame of data
"""
cols = [
"Team", "Season", "Game.ID", "Date", "TOI", 'GF', 'FF', 'xGF', 'CF',
'wshF'
]
df = pd.read_csv("../projection_data/teams/teams_pp.csv")
df = helpers.fix_team(df, "Team")
df = df.sort_values(by=['Season', 'Game.ID', 'Team'])
df['wshF'] = ((df['CF'] - df['GF']) * .2 + df['GF'])
df = df[cols]
df = df.rename(index=str, columns={col: col + "_pp" for col in cols[4:]})
return df
def get_all_sits_data():
"""
Process the All Situations DataFrame
:return: Processed All Situations DataFrame
"""
cols = [
'player', 'player_id', 'season', 'game_id', 'date', 'position', 'team',
'opponent', 'venue', 'toi_on', 'goals', 'a1', 'a2', 'icors', 'iblocks',
'pend', 'pent', 'ifac_win', 'ifac_loss', 'games'
]
# Get All Data
dfs = []
for season in range(2007, 2018):
for pos in ['forwards', 'defensemen']:
print("all", pos, season)
with open("../projection_data/skaters/{}_all_sits_{}.json".format(
pos, season)) as file:
dfs.append(pd.DataFrame(json.load(file)['data'], columns=cols))
# Combine All Forward and Defensemen Data
df = pd.concat(dfs)
# Fix the F*****g Names!!!!!
for team_col in ['team', "opponent", "venue"]:
df = helpers.fix_team(df, team_col)
# Idk
df['toi_on'] = df['toi_on'].astype(float)
# Change over some names for merging
df = df.rename(index=str, columns={"toi_on": "toi_on_all"})
# Get the correct game_id
df['game_id'] = df.apply(
lambda x: str(x['season']) + "0" + str(x['game_id']), axis=1)
return df
def get_all_sits_data():
"""
Get and prepare the All Situations data
:return: DataFrame of data
"""
cols = [
"Team", "Season", "Game.ID", "Date", "Opponent", "Venue", "TOI_all",
"PENT_all", "PEND_all"
]
df = pd.read_csv("./data/teams_all_sits.csv")
df = df.sort_values(by=['Season', 'Game.ID', 'Team'])
for team_col in ['Team', "Opponent", "Venue"]:
df = helpers.fix_team(df, team_col)
df = df.rename(index=str,
columns={
"TOI": "TOI_all",
"PENT": "PENT_all",
"PEND": "PEND_all"
})
return df[cols]