def evaluate(model, df, target_label='Decision', task='test'):
"""
Parameters:
model (built chefboost model): you should pass the return of fit function
df (pandas data frame): data frame you would like to evaluate
task (string): optionally you can pass this train, validation or test
"""
#--------------------------
if target_label != 'Decision':
df = df.rename(columns={target_label: 'Decision'})
#if target is not the last column
if df.columns[-1] != 'Decision':
new_column_order = df.columns.drop('Decision').tolist() + ['Decision']
print(new_column_order)
df = df[new_column_order]
#--------------------------
functions.bulk_prediction(df, model)
enableAdaboost = model["config"]["enableAdaboost"]
if enableAdaboost == True:
df['Decision'] = df['Decision'].astype(str)
df['Prediction'] = df['Prediction'].astype(str)
eval.evaluate(df, task=task)
def evaluate(model, df, task = 'test'): functions.bulk_prediction(df, model) enableAdaboost = model["config"]["enableAdaboost"] if enableAdaboost == True: df['Decision'] = df['Decision'].astype(str) df['Prediction'] = df['Prediction'].astype(str) eval.evaluate(df, task = task)
def evaluate(model, df, task='test'):
"""
Parameters:
model (built chefboost model): you should pass the return of fit function
df (pandas data frame): data frame you would like to evaluate
task (string): optionally you can pass this train, validation or test
"""
functions.bulk_prediction(df, model)
enableAdaboost = model["config"]["enableAdaboost"]
if enableAdaboost == True:
df['Decision'] = df['Decision'].astype(str)
df['Prediction'] = df['Prediction'].astype(str)
eval.evaluate(df, task=task)