代码优化

1 代码评估

1.1 timeit

使用timeit评估一个函数的的施行效率（IPython）



In [1]:

    
import numpy as np
a = np.arange(1000)
%timeit a**2









    



The slowest run took 131.59 times longer than the fastest. This could mean that an intermediate result is being cached.
1000000 loops, best of 3: 1.64 µs per loop



In [2]:

    
%timeit a**2.1









    



The slowest run took 4.59 times longer than the fastest. This could mean that an intermediate result is being cached.
10000 loops, best of 3: 32.7 µs per loop



In [3]:

    
%timeit a*a









    



The slowest run took 62.51 times longer than the fastest. This could mean that an intermediate result is being cached.
100000 loops, best of 3: 1.59 µs per loop

1.2 Profiler



In [4]:

    
% run -t ica.py









    



IPython CPU timings (estimated):
  User   :       6.45 s.
  System :       0.20 s.
Wall time:       3.80 s.



In [6]:

    
% run -p ica.py

1.3 Line-Profiler



In [12]:

    
import profile
from scipy import linalg
def test():
    data = np.random.random((5000,100))
    u,s,v = linalg.svd(data)
    pca = np.dot(u[:,:10].T,data)









    



1 loop, best of 3: 3.29 s per loop

2 代码跑的更快

2.1 使用更好的算法



In [13]:

    
data = np.random.random((5000,100))
%timeit linalg.svd(data)









    



1 loop, best of 3: 3.23 s per loop



In [15]:

    
%timeit linalg.svd(data,full_matrices=False)









    



10 loops, best of 3: 20.3 ms per loop

2.2 数值计算



In [18]:

    
a = np.zeros(1e7)
%timeit global a; a =0*a









    



/Users/gaufung/Anaconda/anaconda/lib/python2.7/site-packages/ipykernel/__main__.py:1: VisibleDeprecationWarning: using a non-integer number instead of an integer will result in an error in the future
  if __name__ == '__main__':






    



10 loops, best of 3: 28.9 ms per loop



In [19]:

    
a = np.zeros(1e7)
%timeit global a; a *= 0









    



/Users/gaufung/Anaconda/anaconda/lib/python2.7/site-packages/ipykernel/__main__.py:1: VisibleDeprecationWarning: using a non-integer number instead of an integer will result in an error in the future
  if __name__ == '__main__':






    



100 loops, best of 3: 9.08 ms per loop



In [ ]: