| import json | |
| import os | |
| import csv | |
| # 定义文件所在的目录 | |
| input_dir = 'E:\python-testn\pythonProject3\hh_1\evaluate_result' | |
| # 获取目录中的所有文件 | |
| files = os.listdir(input_dir) | |
| # with open("token_counts_QS.csv","w", newline='') as csvfile: | |
| # writer = csv.writer(csvfile) | |
| # writer.writerow(["Model", "token_subset_1", "token_subset_2","token_subset_3","token_subset_4"]) | |
| with open("line_counts_QS.csv","w", newline='') as csvfile: | |
| writer = csv.writer(csvfile) | |
| writer.writerow(["Model", "line_subset_1", "line_subset_2","line_subset_3","line_subset_4"]) | |
| # | |
| # with open("CC_QS.csv","w", newline='') as csvfile: | |
| # writer = csv.writer(csvfile) | |
| # writer.writerow(["Model", "CC_subset_1", "CC_subset_2","CC_subset_3","CC_subset_4"]) | |
| for file_name in files: | |
| # 构建完整的文件路径 | |
| input_file_path = os.path.join(input_dir, file_name) | |
| first_underscore_index = file_name.find('_') | |
| # 找到最后一个 - 的位置 | |
| last_dash_index = file_name.rfind('-') | |
| model_name = file_name[first_underscore_index + 1:last_dash_index] | |
| print(model_name) | |
| with open(input_file_path,"r",encoding="utf-8") as file: | |
| data1=json.load(file) | |
| with open("QS.json","r",encoding="utf-8") as file: | |
| data2=json.load(file) | |
| sum0=0 | |
| count0=0 | |
| sum1=0 | |
| count1=0 | |
| sum2=0 | |
| count2=0 | |
| sum3 = 0 | |
| count3 = 0 | |
| for (item1,item2) in zip(data1["humaneval"]["pass@1"],data2): | |
| # #按照token个数划分后的评估结果 | |
| # if item2["token_diff"] == 0: | |
| # index, value = item1 | |
| # print(item2["token_diff"],index,value) | |
| # sum0=sum0+value | |
| # count0=count0+1 | |
| # if item2["token_diff"] == 1: | |
| # index, value = item1 | |
| # print(item2["token_diff"], index, value) | |
| # sum1=sum1+value | |
| # count1=count1+1 | |
| # if item2["token_diff"] == 2: | |
| # index, value = item1 | |
| # print(item2["token_diff"], index, value) | |
| # sum2=sum2+value | |
| # count2=count2+1 | |
| # if item2["token_diff"] == 3: | |
| # index, value = item1 | |
| # print(item2["token_diff"], index, value) | |
| # sum3=sum3+value | |
| # count3=count3+1 | |
| #按照行数划分后的评估结果 | |
| if item2["line_diff"] == 0: | |
| index, value = item1 | |
| print(item2["line_diff"],index,value) | |
| sum0=sum0+value | |
| count0=count0+1 | |
| if item2["line_diff"] == 1: | |
| index, value = item1 | |
| print(item2["line_diff"], index, value) | |
| sum1=sum1+value | |
| count1=count1+1 | |
| if item2["line_diff"] == 2: | |
| index, value = item1 | |
| print(item2["line_diff"], index, value) | |
| sum2=sum2+value | |
| count2=count2+1 | |
| if item2["line_diff"] == 3: | |
| index, value = item1 | |
| print(item2["line_diff"], index, value) | |
| sum3=sum3+value | |
| count3=count3+1 | |
| #按照圈复杂度划分后的评估结果 | |
| # if item2["CC_diff"] == 0: | |
| # index, value = item1 | |
| # print(item2["CC_diff"],index,value) | |
| # sum0=sum0+value | |
| # count0=count0+1 | |
| # if item2["CC_diff"] == 1: | |
| # index, value = item1 | |
| # print(item2["CC_diff"], index, value) | |
| # sum1=sum1+value | |
| # count1=count1+1 | |
| # if item2["CC_diff"] == 2: | |
| # index, value = item1 | |
| # print(item2["CC_diff"], index, value) | |
| # sum2=sum2+value | |
| # count2=count2+1 | |
| # if item2["CC_diff"] == 3 : | |
| # index, value = item1 | |
| # print(item2["CC_diff"], index, value) | |
| # sum3=sum3+value | |
| # count3=count3+1 | |
| mean0=round(sum0/count0*100,2) | |
| mean1=round(sum1/count1*100,2) | |
| mean2=round(sum2/count2*100,2) | |
| mean3=round(sum3/count3*100,2) | |
| print("count_result!!") | |
| print(count0,count1,count2,count3) | |
| print(mean0,mean1,mean2,mean3) | |
| # with open("token_counts_QS.csv", mode='a', newline='', encoding='utf-8') as file: | |
| # writer = csv.writer(file) | |
| # writer.writerow([model_name,mean0,mean1,mean2]) | |
| with open("line_counts_QS.csv", mode='a', newline='', encoding='utf-8') as file: | |
| writer = csv.writer(file) | |
| writer.writerow([model_name,mean0,mean1,mean2,mean3]) | |
| # with open("CC_QS.csv", mode='a', newline='', encoding='utf-8') as file: | |
| # writer = csv.writer(file) | |
| # writer.writerow([model_name,mean0,mean1,mean2,mean3]) | |