KDD CUP99 数据挖掘（1）——数据读取,将txt存为cs

最近在处理KDDcup99的数据，将自己遇到的问题和方法记录下来，以分享给大家。

资源整合

KDD CUP1999的数据集下载地址 http://kdd.ics.uci.edu/databases/kddcup99/下载地址

KDD CUP1999的数据集的介绍  KDD CUP 99数据集

KDD CUP1999 的数据集参考项目下载 可供参考项目地址，下载代码可运行

WEKA学习PPT https://pan.baidu.com/s/1slTz5Bf学习文件

数据下载

下载的KDDCUP99的数据文件是这样的

kddcup.namesAlist of features.

kddcup.data.gzThefull data set (18M; 743M Uncompressed)

kddcup.data_10_percent.gzA10% subset. (2.1M; 75M Uncompressed)

kddcup.newtestdata_10_percent_unlabeled.gz(1.4M; 45M Uncompressed)

kddcup.testdata.unlabeled.gz(11.2M;430M Uncompressed)

kddcup.testdata.unlabeled_10_percent.gz(1.4M;45M Uncompressed)

corrected.gzTestdata with corrected labels.

training_attack_typesAlist of intrusion types.

数据集的介绍请看链接1，把corrected.data文件作为训练集，kddcup.data_10_percent作为测试集即可。

数据读取

下载的文本是纯文本文件，用NotePad++打开另存为.txt文件，方便python读取。下面我做的工作就是添加标签，然后把txt文件另存为csv文件

纯文本文件

添加标识，标识为连接1的文章所示，python代码

---

import pandas as pd

col_names = ["duration","protocol_type","service","flag","src_bytes",

 "dst_bytes","land","wrong_fragment","urgent","hot","num_failed_logins",

 "logged_in","num_compromised","root_shell","su_attempted","num_root",

"num_file_creations","num_shells","num_access_files","num_outbound_cmds",

"is_host_login","is_guest_login","count","srv_count","serror_rate",

 "srv_serror_rate","rerror_rate","srv_rerror_rate","same_srv_rate",

"diff_srv_rate","srv_diff_host_rate","dst_host_count","dst_host_srv_count",

 "dst_host_same_srv_rate","dst_host_diff_srv_rate","dst_host_same_src_port_rate",

 "dst_host_srv_diff_host_rate","dst_host_serror_rate","dst_host_srv_serror_rate",

"dst_host_rerror_rate","dst_host_srv_rerror_rate","label"]  #42个标识

data = pd.read_table("corrected.txt",header=None, sep=',',names = col_names)

print(data.head(10))       #查看前10行

data.to_csv("corrected.csv")    #另存为csv文件

---

提前用excel创建空的corrected.csv文件,要不然报错文件不存在（路径都是绝对路径）。

添加标识后的csv文件

下一步数据预处理。。。。。请看下期。