给出以下最小的可重现示例:
import numpy as np
from numba import jit
# variable number of dimensions
n_t = 8
# q is just a partition of n
q_ddl = 2
n_ddl = 3
np.random.seed(42)
df = np.random.rand(q_ddl*n_t,q_ddl*n_t)
# index array
# ddl_nl is a set of np.arange(n_ddl), ex: [0,1] ; [0,2] or even [0] ...
ddl_nl = np.array([0,1])
ij = np.asarray(np.meshgrid(ddl_nl,ddl_nl,indexing='ij'))
@jit(nopython=True)
def foo(df,ij):
out = np.zeros((n_t,n_ddl,n_ddl))
for i in range(0,n_t):
d_i = np.zeros((n_ddl,n_ddl))
# (q_ddl,q_ddl) non zero values into (n_ddl,n_ddl) shape
d_i[ij[0], ij[1]] = df[i::n_t,i::n_t]
# to check possible solutions
out[i,...] = d_i
return out
out_foo = foo(df,ij)
在禁用foo
的情况下,函数@jit(nopython=True)
正常工作,但是在启用时会引发以下错误:
TypeError: unsupported array index type array(int64, 2d, C) in UniTuple(array(int64, 2d, C) x 2)
发生在广播操作d_i[ij[0], ij[1]] = df[i::n_t,i::n_t]
期间。然后,我确实尝试使用ij
之类的东西将2d索引数组d_i[ij[0].ravel(), ij[1].ravel()] = df[i::n_t,i::n_t].ravel()
展平,这给了我相同的输出,但是又出现了另一个错误:
NotImplementedError: only one advanced index supported
所以我终于尝试通过使用经典的2嵌套for
循环结构来回避这个问题:
tmp = df[i::n_t,i::n_t]
for k,r in enumerate(ddl_nl):
for l,c in enumerate(ddl_nl):
d_i[r,c] = tmp[k,l]
正在启用装饰器的情况下,并给出了预期的结果。
但是我不能停止思考我是否在这里缺少此numpy 2d阵列广播操作的任何与numba兼容的替代方案?任何帮助将不胜感激。
答案 0 :(得分:1)
检查一些值:
In [446]: ddl_nl = np.array([0,1])
...: ij = np.asarray(np.meshgrid(ddl_nl,ddl_nl,indexing='ij'))
...:
In [447]: ij
Out[447]:
array([[[0, 0],
[1, 1]],
[[0, 1],
[0, 1]]])
In [448]: n_t = 8
...: q_ddl = 2
...: n_ddl = 3
In [449]: d_i = np.zeros((n_ddl,n_ddl))
In [450]: d_i
Out[450]:
array([[0., 0., 0.],
[0., 0., 0.],
[0., 0., 0.]])
In [451]: d_i[ij[0], ij[1]]
Out[451]:
array([[0., 0.],
[0., 0.]])
尝试进一步诊断d_i
:
In [452]: d_i = np.arange(9).reshape(3,3)
In [453]: d_i[ij[0], ij[1]]
Out[453]:
array([[0, 1],
[3, 4]])
In [454]: d_i[:2,:2]
Out[454]:
array([[0, 1],
[3, 4]])
当基本切片可行时,为什么要使用高级索引?
我还没有尝试使用numba
进行此操作,但是可能会有更好的工作机会。就是说,枚举循环可能同样快。我对numba
的经验不足,不能肯定地说。
===
很明显,您执行了numpy
不支持的numba
操作:
In [456]: numba.__version__
Out[456]: '0.43.0'
In [457]: @numba.jit
...: def foo(arr):
...: return arr[[1,2,3],[1,2,3]]
...:
In [458]: foo(np.eye(4))
Out[458]: array([1., 1., 1.])
In [459]: @numba.njit
...: def foo(arr):
...: return arr[[1,2,3],[1,2,3]]
...:
In [460]: foo(np.eye(4))
TypingError: Failed in nopython mode pipeline (step: nopython frontend)
Invalid use of Function(<built-in function getitem>) with argument(s) of type(s): (array(float64, 2d, C), tuple(list(int64) x 2))
这很正常。 numba
并没有声称覆盖Python
或numpy
。
但是使用numba
,我们不必避免迭代。实际上,在替换numpy
不能没有迭代的操作时,这是最好的。
In [465]: @numba.njit
...: def foo(arr):
...: out = np.zeros((3,), arr.dtype)
...: for n, (i,j) in enumerate(zip([1,2,3],[1,2,3])):
...: out[n] = arr[i,j]
...: return out
In [466]: foo(np.eye(4))
Out[466]: array([1., 1., 1.])
In [467]: timeit foo(np.eye(4))
6.85 µs ± 28.1 ns per loop (mean ± std. dev. of 7 runs, 100000 loops each)
In [468]: np.eye(4)[[1,2,3],[1,2,3]]
Out[468]: array([1., 1., 1.])
In [469]: timeit np.eye(4)[[1,2,3],[1,2,3]]
13.3 µs ± 31.5 ns per loop (mean ± std. dev. of 7 runs, 100000 loops each)
答案 1 :(得分:1)
还避免使用全局变量(它们在编译时进行了硬编码),并使您的代码尽可能简单(简单意味着只有一个露水循环,如果/否则,...)。如果ddl_nl
数组实际上仅使用np.arange构造,则甚至根本不需要此数组。
示例
import numpy as np
from numba import jit
@jit(nopython=True)
def foo_nb(df,n_ddl,n_t,ddl_nl):
out = np.zeros((n_t,n_ddl,n_ddl))
for i in range(0,n_t):
for ii in range(ddl_nl.shape[0]):
ind_1=ddl_nl[ii]
for jj in range(ddl_nl.shape[0]):
ind_2=ddl_nl[jj]
out[i,ind_1,ind_2] = df[i+ii*n_t,i+jj*n_t]
return out
时间
#Testing and compilation
A=foo(df,ij)
B=foo_nb(df,n_ddl,n_t,ddl_nl)
print(np.allclose(A,B))
#True
%timeit foo(df,ij)
#16.8 µs ± 107 ns per loop (mean ± std. dev. of 7 runs, 100000 loops each)
%timeit foo_nb(df,n_ddl,n_t,ddl_nl)
#674 ns ± 2.56 ns per loop (mean ± std. dev. of 7 runs, 1000000 loops each)