lzddzh / dataminingcompetitionfirstprize Goto Github PK

你好啊，我在运行你们demo的时候发现，在process new feature 运行refresh_train.sh的时候，找不到tool.py这个脚本，想问一下，这个脚本是做什么样的处理的？可以提供一下这个脚本吗？

lines = open(root_loc).readlines()
for line in lines[:500]:
temps = line.strip("\n").split("$")
temps 读取出来的格式是 ['1006,"POS消费","地点551","淋浴","2013/09/01 00:00:32","0.5","124.9"']

'stuId':int(temps[0]) 不太对！

for i in range(1, len(temps)):
records = temps[i].split(",")

len(temps) 就为1 感觉不太对

如果第一步line 换成逗号分隔符（原始数据的分隔符是逗号）
for line in lines[:500]:
temps = line.strip("\n").split(",")

temps 读取出来的格式是 ['1006', '"POS消费"', '"地点551"', '"淋浴"', '"2013/09/01 00:00:32"', '"0.5"', '"124.9"']

for i in range(1, len(temps)):
records = temps[i].split(",")
cate = records[0].strip(""")
location = records[1].strip(""")[2:]
.....

一条记录遍历 6次 感觉也不太对。
看代码的意思是想把一条用户对应的多条消费记录取出来，然后在多条记录里面进行循环汇总，
不知道我有没有理解错？还是code有点小问题，中间少了步骤吗？Python菜鸟，求大神解释，先谢啦！

In readme, you say that after running processDorm, processLibrary ..., the new file will occur in trainProcessed or in testProcessed. The next step is running the refresh_train.sh. However, in the sh file,
it need

python tool.py add Card2Processed.txt normalization=$1 train
python tool.py add Card3Processed.txt normalization=$1 train
python tool.py add FacultyProcessed.txt normalization=$1 train
python tool.py add RankProcessed.txt normalization=$1 train

I think that must miss some file.

在python consume_rank_feature.py这个程序中
extract_rank_feature("../original_data/raw_data/test/subsidy_final_test.txt", final_rank,
score_dict, False)
extract_rank_feature("../original_data/raw_data/train/subsidy_train.txt", final_rank,
score_dict, True)

我跑程序时发生错误：IOError: [Errno 2] No such file or directory: '../original_data/raw_data/test/subsidy_final_test.txt'
数据中不是没有这个文件吗