a=[1,2,3]
b=[1,2,3]
a == b # True
a is b # False
a=func(...) # a is False
if a is False:
# do something
在 a is False這一行竟然結果為False而非True，必須將判斷式改為a==False才會為True。
cimport numpy as cnp
ctypedef cnp.float64_t FLOAT_t
@cython.wraparound(False)
@cython.boundscheck(False)
def myfunc(cnp.ndarray[FLOAT_t, ndim=2] f):
pass
c = Client()
dview = c[:]
asyncs = [dview.map_async(f, [arg]) for arg in args]
while asyncs:
for async in asyncs[:]:
if async.ready():
asyncs.remove(async)
print async.result[0]rc = Client()
rc.direct_view().use_dill()
lv = rc.load_balanced_view()from django.db import reset_queries reset_queries()
來清除query object cache, 使用後，記憶體的使用量就變正常了。
n = 1000
c = 20
a = np.random.rand(n,n)
a[np.diag_indices_from(a)] /= c # 119 microseconds
a.flat[::n+1] /= c # 25.3 microseconds #取出off diagonal的方法 np.delete(a, a.ravel()[::n+1])liblapack.so.3gf to/usr/lib/lapack/liblapack.so.3gf
result = 1
for i in xrange(k):
result = result * A
import theano
import theano.tensor as T
theano.config.warn.subtensor_merge_bug = False
k = T.iscalar("k")
A = T.vector("A")
def inner_fct(prior_result, A):
return prior_result * A
# Symbolic description of the result
result, updates = theano.scan(fn=inner_fct,
outputs_info=T.ones_like(A),
non_sequences=A, n_steps=k)
'''
Scan has provided us with A ** 1 through A ** k.
Keep only the last value. Scan notices this and
does not waste memory saving them.
'''
final_result = result[-1]
power = theano.function(inputs=[A, k], outputs=final_result,
updates=updates)
print power(range(10),2)
#[ 0. 1. 4. 9. 16. 25. 36. 49. 64. 81.]
from theano import shared
state = shared(0)
inc = T.iscalar('inc')
accumulator = function([inc], state;
updates=[(state, state+inc)])
state.get_value() #程式尚未執行，array(0) accumulator(1) #array(0)->array(1) state.get_value() #array(1) accumulator(300) #array(1)->array(301) state.get_value() #array(301) #reset shared variable state.set_value(-1) accumulator(3) state.get_value() #array(-1)->array(2)
decrementor = function([inc], state, updates=[(state, state-inc)]) decrementor(2) state.get_value() #array(2)->array(0)
fn_of_state = state * 2 + inc
# the type (lscalar) must match the shared ariable we
# are replacing with the ``givens`` list
foo = T.lscalar()
skip_shared = function([inc, foo], fn_of_state,
givens=[(state, foo)])
skip_shared(1, 3) # we're using 3 for the state, not state.value
state.get_value() # old state still there, but we didn't use it
#array(0)
簡單的求eigenvalues範例:
剛剛在搜尋資料的時候看到這一篇文章Using Intel MKL in your Python program，裡面說明了如何在python當中使用MKL來做數值計算，兩者可以一起使用放合體技，真是太棒了。
而在Intel MKL link advisor這篇文章中，使用者可以依照自已的需求，填完表格後，就會產生link時所需要的參數，可以直接使用或是寫在Makefile中，也是很方便的做法。
在Compiling 64-bit R 2.10.1 with MKL in Linux中，提到了如何使用MKL來compile R，這對於加速數學計算是很有幫助的。
一些實用的小技巧，可減少多餘的動作
#Python建議使用
x, y= y, x
#建議寫法
print d.get('key', 'not find')
#建議寫法
mval, mpos = min([ (s[i], i) for i in xrange(len(s)) ])
#建議寫法1
text = string.join([ line for line in open('text.txt')], '']
#建議寫法2
text = ''.join([ line for line in open('text.txt')])
#建議寫法3
text = file('text.txt').read()
#建議寫法1, 但後面式子為None會有問題
print (n >= 0) and 'positive' or 'negitive'
#建議寫法2, 解決了None的問題
print (n >= 0 and ['positive'] or ['negitive])[0]
#建議寫法3
print ('negitive', 'positive')[n >= 0]
#建議寫法
d.setdefault(y, { })[x] = 3
#如果為list時
d.setdefault(key, []).append(val)