如何用python解决这个问题

rocky1972

运行一个程序，会打印出如下结果：

1. STREAM version $Revision: 5.9 $
   
2. -------------------------------------------------------------
   
3. This system uses 8 bytes per DOUBLE PRECISION word.
   
4. -------------------------------------------------------------
   
5. Array size = 2097152, Offset = 0
   
6. Total memory required = 48.0 MB.
   
7. Each test is run 10 times, but only
   
8. the *best* time for each is used.
   
9. -------------------------------------------------------------
   
10. Printing one line per active thread....
    
11. -------------------------------------------------------------
    
12. Your clock granularity/precision appears to be 3 microseconds.
    
13. Each test below will take on the order of 319658 microseconds.
    
14.    (= 106552 clock ticks)
    
15. Increase the size of the arrays if this shows that
    
16. you are not getting at least 20 clock ticks per test.
    
17. -------------------------------------------------------------
    
18. WARNING -- The above is only a rough guideline.
    
19. For best results, please be sure you know the
    
20. precision of your system timer.
    
21. -------------------------------------------------------------
    
22. Function      Rate (MB/s)   Avg time     Min time     Max time
    
23. Copy:         268.7902       0.1252       0.1248       0.1258
    
24. Scale:         87.0483       0.3866       0.3855       0.3882
    
25. Add:          109.6098       0.4727       0.4592       0.4977
    
26. Triad:         62.7243       0.8045       0.8024       0.8069
    
27. -------------------------------------------------------------

复制代码

我想把这个程序运行 100遍， 然后统计 Copy/Scale/Add/Triad 的平均值，
刚用python 两天，如果问题弱智，请大家理解。

babykick

首先把你需要的字段通过python抓出来，累加再求平均，看看python基础的文件读写，re模块，你就知道了

BangBull

1. 
   
2. >>> import re
   
3. >>> f = open('a')
   
4. >>> a=f.read()
   
5. >>> copy = re.findall('Copy:.*',a)
   
6. 
   
7. >>> Rate = 0
   
8. >>> Avg_time = 0
   
9. >>> Min_time = 0
   
10. >>> Max_time = 0
    
11. 
    
12. >>> for line in copy:
    
13. ...     line = line.split(':')
    
14. ...     value = re.findall('\S+',line[1].strip())
    
15. ...     Rate += float(value[0])
    
16. ...     Avg_time += float(value[1])
    
17. ...     Min_time += float(value[2])
    
18. ...     Max_time += float(value[3])
    
19. 
    
20. >>> print Rate/len(copy), Avg_time/len(copy), Min_time/len(copy), Max_time/len(copy),                 
    
21. 268.7902 0.1252 0.1248 0.1258
    

复制代码

rocky1972

回复 3# BangBull


    谢谢这位大哥，我用了一个笨办法，cat file | grep Copy | awk `{print $2}` > newfile
    对python确实不熟，惭愧了