# ### 脚本结构介绍 # 每个脚本都主要由两个函数构成:`validate`函数和`score`函数。 import sys import pandas as pd import numpy as np from sklearn.metrics import roc_curve,auc def validate(submit_file, test_file, options=True): submission = pd.read_csv(submit_file,sep =',',dtype={'Prediction':str}) ground_truth = pd.read_csv(test_file,sep=',') #### duplicate ID if len(np.unique(submission['ID'])) != len(submission['ID']) : raise Exception('Duplicate ID in answer File!') #### ID not matched (the length of the ID column of the two files should be the same) if len(np.intersect1d(submission['ID'],ground_truth['ID'])) != len(ground_truth['ID']): raise Exception('ID column of two files not matched!') #### prediction should be decimal var = submission['Prediction'].apply(lambda x : x.isdigit()).any() if var == True: raise Exception('Prediction column only accepts float number!') return { 'code':0, 'message':'validation success' } if __name__ == '__main__': submit_file = sys.argv[1] test_file = sys.argv[2] print(validate(submit_file, test_file,options=True))