def train_model(args):
"""Orchestrates the training of model from commandline arguments."""
with open(args.config, "r") as f:
config = yaml.load(f, Loader=yaml.SafeLoader)
atp_data = read_raw(**config['train_model']['read_features'])
atp_matches_p = choose_features(atp_data,
**config['train_model']['choose_features'])
setFeatureType(atp_matches_p, **config['train_model']['setFeatureType'])
data = split_train_test(atp_matches_p,
**config['train_model']['split_train_test'])
train_features = data[0]
test_features = data[1]
train_labels = data[2]
test_labels = data[3]
np.savetxt(args.savedatapath + "//train_features.csv",
train_features,
fmt='%5s',
delimiter=",")
np.savetxt(args.savedatapath + "//test_features.csv",
test_features,
fmt='%5s',
delimiter=",")
np.savetxt(args.savedatapath + "//train_labels.csv",
train_labels,
fmt='%s',
delimiter=",")
np.savetxt(args.savedatapath + "//test_labels.csv",
test_labels,
fmt='%s',
delimiter=",")
newmodel = fit_xgboost(train_features, train_labels,
**config['train_model']['fit_xgboost'])
f.close()
return newmodel
def run_features(args):
"""Orchestrates the generating of features from commandline arguments."""
with open(args.config, "r") as f:
config = yaml.load(f, Loader=yaml.BaseLoader)
df = read_raw(**config["run_features"]['read_main'])
data = select_columns(df, **config["run_features"]['select_columns'])
h2h_record = read_raw(**config["run_features"]['read_h2h'])
surface_record = read_raw(**config["run_features"]['read_surface'])
df_h2h = add_h2h(data, h2h_record)
df_allfeatures = add_surface_winpct(df_h2h, surface_record)
matches = flip_records(df_allfeatures,
**config["run_features"]['flip_records'])
save_dataset(matches, **config["run_features"]['save_dataset'])
f.close()
return matches
def run_h2h_record(args):
"""Orchestrates the generating of h2h records table from commandline arguments."""
with open(args.config, "r") as f:
config = yaml.load(f, Loader=yaml.BaseLoader)
df = read_raw(**config["run_h2h_record"]['read_raw'])
h2h_record = calculate_h2h(df)
save_dataset(h2h_record, **config["run_h2h_record"]['save_dataset'])
f.close()
return h2h_record
def run_rankingstable(args):
"""Orchestrates the generation of rankings table from commandline arguments."""
with open(args.config, "r") as f:
config = yaml.load(f, Loader=yaml.BaseLoader)
df = read_raw(**config["run_rankingstable"]['read_raw'])
srank = gen_rankings_static(df)
save_dataset(srank, **config["run_rankingstable"]['save_dataset'])
f.close()
return srank
def run_trimdata(args):
"""Orchestrates the trim data functionalities from commandline arguments."""
with open(args.config, "r") as f:
config = yaml.load(f, Loader=yaml.BaseLoader)
df = read_raw(**config["run_trimdata"]['read_raw'])
df_trim = trim_columns(df, **config["run_trimdata"]['trim_columns'])
save_dataset(df_trim, **config["run_trimdata"]['save_dataset'])
f.close()
return df_trim
def df_to_db(args):
"""Orchestrates the writing of csv files to database from commandline arguments."""
with open(args.config, "r") as f:
config = yaml.load(f, Loader=yaml.BaseLoader)
if args.rds == 'True': # Setting up the database in RDS
engine_string = conf.rds_engine_string
else: # Setting up the database in local using sqlite
engine_string = conf.local_engine_string
if args.option in ['Ranking', 'H2H', 'SurfaceWinPct']:
engine = sql.create_engine(engine_string, echo=True)
Base.metadata.create_all(engine)
df = read_raw(**config['create_db_local'][args.option]['read_main'])
setFeatureType(
df, **config['create_db_local'][args.option]['setFeatureType'])
df.to_sql(args.option, con=engine, if_exists='replace', index=False)
logger.info("Database created for %s", args.option)
else:
raise logger.error("%s is not a valid table option", args.option)