Hi numpy forum
I need to efficiently handle some large (300 MB) recordlike binary
files, where some data fields are less than a byte and thus cannot be
mapped in a record dtype immediately.
I would like to be able to access these derived arrays in a memory
efficient manner but I cannot figure
Announcing Numexpr 1.1
Numexpr is a fast numerical expression evaluator for NumPy. With it,
expressions that operate on arrays (like 3*a+4*b) are accelerated
and use less memory than doing the same calculation in Python.
The expected speed-ups
On Thu, Jan 15, 2009 at 11:53 PM, David Cournapeau courn...@gmail.comwrote:
On Fri, Jan 16, 2009 at 12:18 AM, Darren Dale dsdal...@gmail.com wrote:
Hi Jarrod,
On Wed, Jan 14, 2009 at 2:21 AM, Jarrod Millman mill...@berkeley.edu
wrote:
Due to the fact that I was tired of adding
On Thu, January 15, 2009 6:17 pm, Sturla Molden wrote:
Is it possible to make f2py raise an exception if a fortran routine
signals an error?
If I e.g. have
subroutine foobar(a, ierr)
Can I get an exception automatically raised if ierr != 0?
Yes, for that you need to provide your own
Hello.
I am terribly sorry. I was mistaken last night. I had the latest Matplotlib
version 0.98.5.2 and I thought the bug was fixed but it hasn't. Let me explain.
In the file MPL_isnan.h line 26 there is a declaration:
typedef long int MPL_Int64
This is fine for
Francesc Alted schrieb:
Numexpr is a fast numerical expression evaluator for NumPy. With it,
expressions that operate on arrays (like 3*a+4*b) are accelerated
and use less memory than doing the same calculation in Python.
The expected speed-ups for Numexpr respect to NumPy are between 0.95x
On Jan 16, 2009, at 10:51 AM, josef.p...@gmail.com wrote:
I have a regression result with masked arrays that produces a masked
array output, estm5.yhat, and I want to test equality to the benchmark
case, estm1.yhat, with the asserts in numpy.testing, but I am getting
strange results.
...
A Friday 16 January 2009, j...@physics.ucf.edu escrigué:
Hi Francesc,
Numexpr is a fast numerical expression evaluator for NumPy. With
it, expressions that operate on arrays (like 3*a+4*b) are
accelerated and use less memory than doing the same calculation in
Python.
Please pardon my
On 1/16/2009 2:16 PM, Pearu Peterson wrote:
Yes, for that you need to provide your own fortran call code
using f2py callstatement construct. The initial fortran call
code can be obtained from f2py generated modulenamemodule.c file,
for instance.
Thank you, Pearu :)
f2py is really a
Hi Francesc,
this is a wonderful project ! I was just wondering if you would /
could support single precision float arrays ?
In 3+D image analysis we generally don't have enough memory to effort
double precision; and we could save our selves lots of extra C coding
(or Cython) coding of we could
On Fri, Jan 16, 2009 at 10:59 AM, Pierre GM pgmdevl...@gmail.com wrote:
On Jan 16, 2009, at 10:51 AM, josef.p...@gmail.com wrote:
I have a regression result with masked arrays that produces a masked
array output, estm5.yhat, and I want to test equality to the benchmark
case, estm1.yhat, with
A Friday 16 January 2009, Gregor Thalhammer escrigué:
I also gave a try to the vector math library (VML), contained in
Intel's Math Kernel Library. This offers a fast implementation of
mathematical functions, operating on array. First I implemented a C
extension, providing new ufuncs. This
A Friday 16 January 2009, Sebastian Haase escrigué:
Hi Francesc,
this is a wonderful project ! I was just wondering if you would /
could support single precision float arrays ?
As I said before, it is doable, but I don't know if I will have time
enough to implement this myself.
In 3+D image
Francesc Alted wrote:
A Friday 16 January 2009, j...@physics.ucf.edu escrigué:
Right
now, I'm not quite sure whether the problem you are solving is merely
the case of expressions-in-strings, and there is no advantage for
expressions-in-code, or whether your expressions-in-strings are
faster
John Hunter wrote:
Andrew, since you are the original author of the isnan port, could you
patch the branch and the trunk to take care of this?
Done in r6791 and r6792.
Sorry for the trouble.
Now I just hope we don't get a problem with long long, although now if
_ISOC99_SOURCE is defined,
Francesc Alted schrieb:
A Friday 16 January 2009, Gregor Thalhammer escrigué:
I also gave a try to the vector math library (VML), contained in
Intel's Math Kernel Library. This offers a fast implementation of
mathematical functions, operating on array. First I implemented a C
extension,
Thanks for all the ideas. I think I will look into the
scikits.delaunay, Rbf, or gaussian smoothing approach. My best idea
is similar to the Gaussian smoothing. Anyway, all of the missing data
gaps seem to be small enough that I expect any of these methods to
accomplish my purpose.
Hi,
I am just visiting colleagues in the Cuban Neuroscience Center, and of
course I'm trying to persuade them that Python and open-source are the
way forward.
This is made more difficult because several projects - for example
pyglet - have their repositories on Google code. Google, unlike any
On Fri, Jan 16, 2009 at 7:07 PM, Matthew Brett matthew.br...@gmail.comwrote:
So, please, if you are considering google code for hosting, consider
other options.
Seems odd that you'd post that from a gmail account. I do sympathize with
your suggestion, but I don't have a better alternative to
On Fri, 16 Jan 2009 19:24:56 -0500
Kevin Jacobs jac...@bioinformed.com bioinfor...@gmail.com wrote:
On Fri, Jan 16, 2009 at 7:07 PM, Matthew Brett matthew.br...@gmail.comwrote:
So, please, if you are considering google code for hosting, consider
other options.
Seems odd that you'd
On Fri, Jan 16, 2009 at 18:07, Matthew Brett matthew.br...@gmail.com wrote:
Hi,
I am just visiting colleagues in the Cuban Neuroscience Center, and of
course I'm trying to persuade them that Python and open-source are the
way forward.
This is made more difficult because several projects -
Suppose I have a 3d array, A, with dimensions 2 x 2 x N, and a 2d 2 x
N array, u. I interpret A as N 2x2 matrices and u as N 2d vectors.
Suppose I want to apply the mth matrix to the mth vector, i.e.
A[, , m] u[, m] = v[, m]
Aside from doing
A[0,0,:] u[0,:] + A[0,1,:] u[1,:] = v[0,:]
and
Note that Apple has a similar library called vForce:
http://developer.apple.com/ReleaseNotes/Performance/RN-vecLib/index.html
http://developer.apple.com/documentation/Performance/Conceptual/vecLib/Reference/reference.html
I think these libraries use several techniques and are not
2009/1/16 Gregor Thalhammer gregor.thalham...@gmail.com:
Francesc Alted schrieb:
Wow, pretty nice speed-ups indeed! In fact I was thinking in including
support for threading in Numexpr (I don't think it would be too
difficult, but let's see). BTW, do you know how VML is able to achieve
a
24 matches
Mail list logo