On 2. apr. 2012, at 15:47, David Cournapeau wrote:
>
>
> On Sun, Apr 1, 2012 at 2:28 PM, Kamesh Krishnamurthy
> wrote:
> Hello all,
>
> I profiled NumPy EIG and MATLAB EIG on the same Macbook pro, and both were
> linking to the Accelerate framework BLAS. NumPy turns out to be ~4x slower.
On Mon, Apr 2, 2012 at 6:18 PM, Aronne Merrelli
wrote:
> On Sun, Apr 1, 2012 at 8:28 AM, Kamesh Krishnamurthy
> wrote:
>> Hello all,
>>
>> I profiled NumPy EIG and MATLAB EIG on the same Macbook pro, and both were
>> linking to the Accelerate framework BLAS. NumPy turns out to be ~4x slower.
>>
On Sun, Apr 1, 2012 at 8:28 AM, Kamesh Krishnamurthy wrote:
> Hello all,
>
> I profiled NumPy EIG and MATLAB EIG on the same Macbook pro, and both were
> linking to the Accelerate framework BLAS. NumPy turns out to be ~4x slower.
> I've posted details on Stackoverflow:
> http://stackoverflow.com/q
Le 2 avril 2012 18:36, Frédéric Bastien a écrit :
> numpy.random are not optimized. If matlab use the random number from
> mkl, they will be much faster.
In that case this is indeed negligible:
In [1]: %timeit np.random.randn(2000, 2000)
1 loops, best of 3: 306 ms per loop
--
Olivier
http://tw
numpy.random are not optimized. If matlab use the random number from
mkl, they will be much faster.
Fred
On Mon, Apr 2, 2012 at 12:04 PM, David Cournapeau wrote:
>
>
> On Mon, Apr 2, 2012 at 4:45 PM, Chris Barker wrote:
>>
>> On Mon, Apr 2, 2012 at 2:25 AM, Nathaniel Smith wrote:
>> > To see i
On Mon, Apr 2, 2012 at 4:45 PM, Chris Barker wrote:
> On Mon, Apr 2, 2012 at 2:25 AM, Nathaniel Smith wrote:
> > To see if this is an effect of numpy using C-order by default instead of
> > Fortran-order, try measuring eig(x.T) instead of eig(x)?
>
> Just to be clear, .T re-arranges the strides
On Mon, Apr 2, 2012 at 2:25 AM, Nathaniel Smith wrote:
> To see if this is an effect of numpy using C-order by default instead of
> Fortran-order, try measuring eig(x.T) instead of eig(x)?
Just to be clear, .T re-arranges the strides (Making it Fortran
order), butyou'll have to make sure your ari
On Sun, Apr 1, 2012 at 2:28 PM, Kamesh Krishnamurthy wrote:
> Hello all,
>
> I profiled NumPy EIG and MATLAB EIG on the same Macbook pro, and both were
> linking to the Accelerate framework BLAS. NumPy turns out to be ~4x slower.
> I've posted details on Stackoverflow:
> http://stackoverflow.com/q
Changing the array to Fortran order using numpy.ndarray.T does not help
much in my machine. But, this may be important since the LAPACK routines
are written in Fortran 90.
On 2 April 2012 12:25, Nathaniel Smith wrote:
> To see if this is an effect of numpy using C-order by default instead of
> F
To see if this is an effect of numpy using C-order by default instead of
Fortran-order, try measuring eig(x.T) instead of eig(x)?
-n
On Apr 1, 2012 2:28 PM, "Kamesh Krishnamurthy" wrote:
> Hello all,
>
> I profiled NumPy EIG and MATLAB EIG on the same Macbook pro, and both were
> linking to the
Hello all,
I profiled NumPy EIG and MATLAB EIG on the same Macbook pro, and both were
linking to the Accelerate framework BLAS. NumPy turns out to be ~4x slower.
I've posted details on Stackoverflow:
http://stackoverflow.com/q/9955021/974568
Can someone please let me know the reason for the perfo
11 matches
Mail list logo
