我正在尝试在IPython中进行文件分析,生成一些分析统计信息输出,然后将其传递给一些Python分析GUI工具,如KCachegrind。这是我的代码试图这样做。所有代码都在IPython中执行。
# example function for profiling
def factorial(n):
if n == 0:
return 1.0
else:
return float(n) * factorial(n-1)
def taylor_sin(n):
res = []
for i in range(n):
if i % 2 == 1:
res.append((-1)**((i-1)/2)/float(factorial(i)))
else:
res.append(0.0)
return res
# generate cProfile output (the stats, not the text)
%prun -D prof.out x = taylor_sin(500)
# do conversion using pyprof2calltree
# to do it in IPython, add prefix '!' to code to make it Shell-command code
!pyprof2calltree -i prof.out -o prof.calltree
现在IPython会输出一条错误消息:
!pyprof2calltree -i prof.out -o prof.calltree
/bin/sh: 1: pyprof2calltree: not found
这是说我还没有pyprof2calltree
添加到环境路径或类似的东西?怎么解决?
我可以在纯shell命令中完美运行它。但我不想经常在IPython和终端之间切换,我想在IPython中做所有的事情。我知道添加前缀!
会使代码像shell命令一样运行,但为什么它会向我提出错误,如上所示?
Jian@Home-PC:~/Dropbox/Coding/Python$ pyprof2calltree -i prof.out -o prof.calltree
writing converted data to: prof.calltree
Jian@Home-PC:~/Dropbox/Coding/Python$
IPython与Anaconda py3.4一起安装;操作系统Ubuntu 14.04;通过pip
安装pyprof2calltree
答案 0 :(得分:1)
来自pyprof2calltree documentation中的ipython
示例:
>>> from pyprof2calltree import convert, visualize
>>> visualize('prof.out')
>>> convert('prof.out', 'prof.calltree')
或者:
>>> results = %prun -r x = taylor_sin(500)
>>> visualize(results)
>>> convert(results, 'prof.calltree')
您也可以尝试:
>>> %run -m pyprof2calltree -i prof.out -o prof.calltree