blaze.zeros() slowness

[blaze]

blaze.zeros() slowness

[FrancescAlted]

If that helps, the slwodown is much less noticeable with multidimensional arrays:

import blaze as blz
import numpy as np
from time import time

shape = (100,100,100)
len_ = np.prod(shape)
print "len for array:", len_

t0 = time()
a = np.arange(len_).reshape(shape)
print "numpy creation time: %.3f" % (time() - t0,)

t0 = time()
b = blz.Array(a, dshape='%d,%d,%d, int32' % shape)
t1 = time() - t0
print "Final datashape:", b.datashape
print "blaze.Array creation time: %.3f" % (t1,)

t0 = time()
c = blz.zeros(dshape='%d,%d,%d, int32'% shape)
t2 = time() - t0
print "Final datashape:", c.datashape
print "blaze.zeros creation time: %.3f" % (t2,)
print "time ratio blaze.Array vs blaze.zeros: %.1fx" % (t2 / t1,)

and the output:

len for array: 1000000
numpy creation time: 0.008
Final datashape: 100, 100, 100, int32
blaze.Array creation time: 0.030
Final datashape: 100, 100, 100, int32
blaze.zeros creation time: 0.818
time ratio blaze.Array vs blaze.zeros: 27.7x

Although the slowdown is still important...

[aterrel]

This looks to be resolved:

In [6]: run blaze_zeros_test.py
len for array: 1000000
numpy creation time: 0.004
Final datashape: 100, 100, 100, int32
blaze.Array creation time: 0.397
Final datashape: 100, 100, 100, int32
blaze.zeros creation time: 0.002
time ratio blaze.Array vs blaze.zeros: 0.0x