Menu
▾
▴
pysparse-users
You can subscribe to this list here.
| 2005 |
Jan
|
Feb
|
Mar
|
Apr
|
May
(1) |
Jun
(3) |
Jul
|
Aug
|
Sep
(1) |
Oct
|
Nov
|
Dec
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2006 |
Jan
|
Feb
|
Mar
(3) |
Apr
|
May
|
Jun
|
Jul
(1) |
Aug
(2) |
Sep
|
Oct
|
Nov
|
Dec
(1) |
| 2007 |
Jan
(2) |
Feb
(5) |
Mar
(2) |
Apr
|
May
(2) |
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| 2008 |
Jan
(2) |
Feb
|
Mar
|
Apr
|
May
|
Jun
(6) |
Jul
(6) |
Aug
|
Sep
(5) |
Oct
|
Nov
|
Dec
(1) |
| 2009 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
(2) |
Aug
(4) |
Sep
(6) |
Oct
(8) |
Nov
(19) |
Dec
(2) |
| 2010 |
Jan
(4) |
Feb
(4) |
Mar
(17) |
Apr
(12) |
May
(10) |
Jun
(17) |
Jul
(2) |
Aug
(5) |
Sep
|
Oct
|
Nov
|
Dec
|
| 2011 |
Jan
|
Feb
|
Mar
(3) |
Apr
(2) |
May
|
Jun
(9) |
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| 2012 |
Jan
(4) |
Feb
(12) |
Mar
|
Apr
|
May
(3) |
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| 2013 |
Jan
|
Feb
|
Mar
|
Apr
(2) |
May
(1) |
Jun
(1) |
Jul
(1) |
Aug
|
Sep
|
Oct
(1) |
Nov
|
Dec
(2) |
| 2014 |
Jan
|
Feb
(4) |
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
(2) |
Nov
|
Dec
(1) |
|
From: Daniel W. <dan...@gm...> - 2008-06-17 15:09:18
|
On Tue, Jun 17, 2008 at 5:33 AM, Artur Palha <art...@gm...> wrote:
> Hi to you all, I hope someone can help me.
>
> I am trying to do the following with pysparse but I am having problems doing it.
>
> I have a dense vector, say:
>
> x = numpy.arange(0,10)
>
> and a sparse matrix:
>
> A = pysparse.spmatrix.ll_mat(10,10)
>
> What I wish to do is:
>
> x = x + A[:, 2]
>
> But this gives the following error:
> TypeError: unsupported operand type(s) for +: 'float' and 'll_mat'
If I remember correctly, ll_mat would need to have the __iter__ method
to work with numpy. numpy checks the type of ll_mat then looks for one
of __getitem__, __iter__, __array__ or __add__. I don't think ll_mat
has these. The code I use has a wrapper which makes pysparse more
pythonic. You may prefer to use the wrapper if it suits your purposes.
<http://matforge.org/fipy/browser/trunk/fipy/tools/pysparseMatrix.py>
>
> Then I tried like this:
>
> x = numpy.arange(0,10)
>
> A = pysparse.sparray.sparray((10,10))
>
> With this it works ok but then, when I try to do a slice:
>
> A_reduced = A[0:5, 0:5]
>
> It gives an error:
> TypeError: unsupported operand type(s) for +: 'int' and 'slice'
Not sure about why this doesn't work. It was probably implemented and
never tested properly. I don't believe it does anything useful
anyway.
> Anybody can give a help?
>
> Also, what is the main difference between spmatrix and sparray?
I would avoid using sparray. You also might want to check out scipy as
their sparse support is much better now.
--
Daniel Wheeler
|
|
From: Artur P. <art...@gm...> - 2008-06-17 09:33:02
|
Hi to you all, I hope someone can help me. I am trying to do the following with pysparse but I am having problems doing it. I have a dense vector, say: x = numpy.arange(0,10) and a sparse matrix: A = pysparse.spmatrix.ll_mat(10,10) What I wish to do is: x = x + A[:, 2] But this gives the following error: TypeError: unsupported operand type(s) for +: 'float' and 'll_mat' Then I tried like this: x = numpy.arange(0,10) A = pysparse.sparray.sparray((10,10)) With this it works ok but then, when I try to do a slice: A_reduced = A[0:5, 0:5] It gives an error: TypeError: unsupported operand type(s) for +: 'int' and 'slice' Anybody can give a help? Also, what is the main difference between spmatrix and sparray? Thanking in advance Best regards -artur palha |
|
From: Daniel W. <dan...@gm...> - 2008-01-29 15:52:48
|
I committed the changes and bumped the version number. I'll hold off
on a release
unless there is a pressing need.
Cheers
On Jan 28, 2008 4:34 PM, Dominique Orban <dom...@gm...> wrote:
> Hello,
>
> I wrote an addition to the UMFPack module of PySparse that lets users
> retrieve the details of the factorization. Here is an example:
>
> from pysparse import spmatrix, umfpack
> A = spmatrix.ll_mat_from_mtx('bcsstk11.mtx')
> LU = umfpack.factorize(A)
> (L, U, P, Q, R, do_recip) = LU.lu()
>
> The help of lu() is as follows:
>
> -------
> lu(...)
> self.lu()
> Returns L and U factors, permutation and scaling information.
>
> Use: (L, U, P, Q, R, do_recip) = self.lu().
> The original matrix A is factorized into
> L U = P R A Q
> where L is unit lower triangular,
> U is upper triangular,
> P and Q are permutation matrices,
> R is a row-scaling diagonal matrix such that
> the i-th row of A has been divided by R[i] if do_recip = True,
> the i-th row of A has been multiplied by R[i] if do_recip = False.
>
> L and U are returned as ll_mat sparse matrices.
> P, Q and R are returned as NumPy arrays.
> -------
>
> I would very much like to see this function included in PySparse. I
> will e-mail the code directly to Daniel Wheeler.
>
> Comments welcome.
>
> Cheers,
> Dominique
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2008.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> Pysparse-users mailing list
> Pys...@li...
> https://lists.sourceforge.net/lists/listinfo/pysparse-users
>
--
Daniel Wheeler
|
|
From: Dominique O. <dom...@gm...> - 2008-01-28 21:34:10
|
Hello,
I wrote an addition to the UMFPack module of PySparse that lets users
retrieve the details of the factorization. Here is an example:
from pysparse import spmatrix, umfpack
A = spmatrix.ll_mat_from_mtx('bcsstk11.mtx')
LU = umfpack.factorize(A)
(L, U, P, Q, R, do_recip) = LU.lu()
The help of lu() is as follows:
-------
lu(...)
self.lu()
Returns L and U factors, permutation and scaling information.
Use: (L, U, P, Q, R, do_recip) = self.lu().
The original matrix A is factorized into
L U = P R A Q
where L is unit lower triangular,
U is upper triangular,
P and Q are permutation matrices,
R is a row-scaling diagonal matrix such that
the i-th row of A has been divided by R[i] if do_recip = True,
the i-th row of A has been multiplied by R[i] if do_recip = False.
L and U are returned as ll_mat sparse matrices.
P, Q and R are returned as NumPy arrays.
-------
I would very much like to see this function included in PySparse. I
will e-mail the code directly to Daniel Wheeler.
Comments welcome.
Cheers,
Dominique
|
|
From: Jonathan G. <gu...@ni...> - 2007-05-21 16:05:07
|
On May 21, 2007, at 10:32 AM, Daniel Wheeler wrote: > On May 4, 2007, at 8:33 PM, Dominique Orban wrote: > >> Not that I have noticed. Since I sent the bug report, I observed >> that the above also affected >> PySparse itself. For instance, the statement >> >> from pysparse import spmatrix, itsolvers, precon > > It was broken for me too, but now it seems to be working. Weird. > Jon, Is this something that you repaired? I doubt it. I don't have checkin privileges for pysparse. |
|
From: Daniel W. <dan...@ni...> - 2007-05-21 14:34:12
|
I just fixed the problem below and made a new release. Please take a
look and check that it works.
If I don't hear any complaints, I'll put together the window binaries
and Jon will do the dmg in the next few weeks.
On May 4, 2007, at 8:33 PM, Dominique Orban wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Daniel Wheeler wrote:
>> On May 4, 2007, at 6:46 PM, Dominique Orban wrote:
>>
>>>
>>> The bug is in spmatrix_api.h, in the "import_spmatrix" macro. I
>>> believe the first line should read
>>>
>>> PyObject *spmatrix = PyImport_ImportModule("pysparse.spmatrix");
>>>
>>> instead of
>>>
>>> PyObject *spmatrix = PyImport_ImportModule("spmatrix");
>>
>> Thanks for the bug report. Are there any other places with similar
>> problems?
>
> Not that I have noticed. Since I sent the bug report, I observed
> that the above also affected
> PySparse itself. For instance, the statement
>
> from pysparse import spmatrix, itsolvers, precon
It was broken for me too, but now it seems to be working. Weird. Jon,
Is this something that you repaired?
>
> caused the error message "no module named spmatrix". I presume the
> error was coming from itsolvers
> and/or precon.
--
Daniel Wheeler
|
|
From: Daniel W. <dan...@ni...> - 2007-03-26 14:51:35
|
Hi, The numpy header file issue has been fixed on HEAD. Please test the fix by rebuilding pysparse if you get a chance and check that it works for you. It turns out that numpy has a get_include() function that gives the location of the header files. Delete any header files that you have copied to locations outside of site-packages such as pysparse/Include before building. If I don't hear any complaints I'll do a release at the end of the week. Cheers On Jan 3, 2007, at 9:37 AM, Roman Geus wrote: >> >> - If you are compiling against numpy, the numpy header files >> don't seem >> to be going into the right place at least on my two computers. >> I haven't resolved this problem yet. For now, I was just >> copying the >> .h files from .../site-packages/python/numpy/core/include/ >> to Include/Lib/numpy >> in my working copy of pysparse. -- Daniel Wheeler |
|
From: Roman G. <rom...@gm...> - 2007-03-02 10:48:43
|
2007/2/28, Daniel Wheeler <dan...@ni...>: > Does anyone else have problems accessing the mailing list archive at > sourceforge? Seems to work for me, at least currently. > I think this list would be better served by archiving at gmane. > <http://www.gmane.org> > Not that I am volunteering to do the set up. Any thoughts? I wouldn't mind, but since these are very low volume lists, I don't feel a strong urge to change them myself ;-) Roman > > > -- > > Daniel Wheeler > > > ------------------------------------------------------------------------- > Take Surveys. Earn Cash. Influence the Future of IT > Join SourceForge.net's Techsay panel and you'll get the chance to share your > opinions on IT & business topics through brief surveys-and earn cash > http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV > _______________________________________________ > Pysparse-users mailing list > Pys...@li... > https://lists.sourceforge.net/lists/listinfo/pysparse-users > > |
|
From: Daniel W. <dan...@ni...> - 2007-02-28 16:45:51
|
Does anyone else have problems accessing the mailing list archive at sourceforge? I think this list would be better served by archiving at gmane. <http://www.gmane.org> Not that I am volunteering to do the set up. Any thoughts? -- Daniel Wheeler |
|
From: Daniel W. <dan...@ni...> - 2007-02-21 16:09:45
|
Hi William,
When I run this I just get an error saying:
>>> A.update_add_mask( A,ind,ind,mask,mask )
Traceback (most recent call last):
File "<stdin>", line 1, in ?
TypeError: a float is required
I don't think "update_add_mask" is particularly useful as a method.
We tend to use
"update_add_at" as it provides more useful functionality.
Cheers
On Feb 15, 2007, at 12:32 PM, William Hunter wrote:
> Daniel/Roman;
>
> Please check out the following behaviour, or else I'm doing
> something wrong:
>
> A = spmatrix.ll_mat( 2,2 )
> B = asarray( [[1,2][3,4]] )
> ind = asarray( [0,1] )
> mask = ones( (2),dtype=int )
> A.update_add_mask( A,ind,ind,mask,mask )
>
> First, am I using it correctly? (I'm sure I am...)
>
> Secondly, one expects an array:
>
> A = B,
>
> but instead get the following:
>
> A = B.T
>
> Of course one can enter the transpose (B.T) as the update array,
> but that's a work around, also,
> if B is symmetric, it doesn't make a difference anyway, it's just
> that it's confusing at first...
>
> This is fairly trivial to fix (*if* it is a minor bug) in the src
> code, I think, haven't looked yet.
>
> So, while I'm on the subject of src code, I notice that the
> docstrings aren't all the same -- can I correct
> the above mentioned assumed 'bug' and change the docstrings while
> I'm at it?
Whatever you think is the most sensible approach.
> I'll also attempt to fix setup.py so that it finds the NumPy
> headers, this however, will not be trivial, unless
> I assume a standard installation (lib/python/site-packages or
> prefix=/usr/local).
Isn't this a problem with numpy? If you think it is a numpy problem I
think we should approach the numpy people
about it. What do you think?
> By the way, still busy with the manual, adding the other methods
> too (e.g. update_add_at)...
Thanks!
>
> Cheers,
> William
>
>
>
--
Daniel Wheeler
|
|
From: Daniel W. <dan...@ni...> - 2007-02-06 15:19:23
|
On Feb 6, 2007, at 3:56 AM, Roman Geus wrote: > Hi Daniel > >> Should I join as a developer or send the strings via the >> developer mailing list? >> >> Unfortunately, I am not the author of pysparse and I don't have >> sufficient >> administrative privileges >> to add you as a developer. If Roman is reading these emails he may >> add you, >> I am not sure. Failing that >> you can just send me your changes and I will try and add them. I am >> forwarding this to the list so Roman >> can see it. > > Since you are the most active pysparse developer and user I have given > you full access to the sourceforge admin interface. I hope that's is > OK with you. Thanks Roman. > > -- Roman -- Daniel Wheeler |
|
From: Roman G. <rom...@gm...> - 2007-02-06 08:56:17
|
Hi Daniel > Should I join as a developer or send the strings via the > developer mailing list? > > Unfortunately, I am not the author of pysparse and I don't have sufficient > administrative privileges > to add you as a developer. If Roman is reading these emails he may add you, > I am not sure. Failing that > you can just send me your changes and I will try and add them. I am > forwarding this to the list so Roman > can see it. Since you are the most active pysparse developer and user I have given you full access to the sourceforge admin interface. I hope that's is OK with you. -- Roman |
|
From: Daniel W. <dan...@ni...> - 2007-02-05 16:06:57
|
Hi William,
Thanks for your interest.
On Feb 5, 2007, at 7:11 AM, Africanis wrote:
>
> Message body follows:
>
> Daniel;
>
> How can I contribute in terms of docstrings?
Docstrings would certainly be useful for the project. Arguably, the
project needs another level
of python wrapping to make the interface more pythonic. We actually
have a wrapper class that
we use for our project:
<http://matdl-osi.org/fipy/browser/trunk/fipy/tools/
sparseMatrix.py>
Anyhow, I guess this should be included with pysparse, but I haven't
got round it.
If you'd like to get involved with that process then go ahead.
Failing that, I think docstrings would be very useful.
> Should I join as a developer or send the strings via the
> developer mailing list?
Unfortunately, I am not the author of pysparse and I don't have
sufficient administrative privileges
to add you as a developer. If Roman is reading these emails he may
add you, I am not sure. Failing that
you can just send me your changes and I will try and add them. I am
forwarding this to the list so Roman
can see it.
Cheers
>
> William
>
> --
> This message has been sent to you, a registered SourceForge.net user,
> by another site user, through the SourceForge.net site. This message
> has been delivered to your SourceForge.net mail alias. You may reply
> to this message using the "Reply" feature of your email client, or
> using the messaging facility of SourceForge.net at:
> https://sourceforge.net/sendmessage.php?touser=1496516
>
--
Daniel Wheeler
|
|
From: Daniel W. <dan...@ni...> - 2007-01-03 19:40:41
|
On Jan 2, 2007, at 3:28 PM, Daniel Wheeler wrote: > The changes for compatibility with numpy have now been merged back > onto HEAD. I will do a new release in the near future. I forgot to notify about two things. Firstly, you will need version 1.0 of numpy. Version 0.9.8 didn't have the correct header files for backwards compatibility with old Numeric modules. Secondly, at the current time, at least for me, the numpy header files do not seem to be going in the right place with the numpy install (same thing has happend on my mac and linux). You'll have to copy or link the header files from python2.4/site- packages/numpy/core/include to include/numpy in the base pysparse directory from where you are building. Cheers > > Cheers > > On Dec 21, 2006, at 12:36 PM, Daniel Wheeler wrote: > >> Numeric is dying slowly and being replaced by numpy. >> With this in mind, I have updated pysparse to use the new numpy, >> this is all on a branch at the moment. >> It was straightforward matter of changing the imported header >> files from Numeric's to numpy's. I am sending this email >> to check to see if everybody is happy with the changes that have >> been made. If there are no objections, I can merge back >> onto the trunk and maybe make a new release. Changes and issues >> are as follows: >> >> - Created a C macro defined from setup.py that either imports >> numpy's or Numeric's headers depending >> on the flag "use_numpy". The C macro was confusing me for >> awhile, but I think it works >> now. >> >> - I have changed the name space of itsolvers, spmatrix, jdsym >> etc to all lie under pysparse. So, instead of >> >> >>> import spmatrix >> >> you do >> >> >>> from pysparse import spmatrix >> >> This change stops site-packages from being polluted. At the >> moment the branch version is using pysparse_numpy >> or pysparse_Numeric as the pysparse name space for testing >> purposes. >> >> - The branch can be checked out via: >> >> $ cvs -d:ext:ui...@py...:/cvsroot/pysparse >> co -r branch-numpy pysparse >> >> - I have updated the Examples and Test directories to import >> from pysparse. They all seem to work apart from stuff that was >> broken already. >> >> - I got rid of pysparse_version.py and replaced it with an >> __init__.py in Lib, which does the imports and holds a __version__ >> variable. >> So you can now do: >> >> >>> import pysparse_numpy >> >>> print pysparse_numpy.__version__ >> 0.34.032-numpy >> >> - I think the default flag should be use_numpy=True since most >> users will now be getting numpy rather than Numeric. This could >> be automated with a try and except, which tests for the >> availability of numpy or Numeric. What do you think? >> >> - You can diff from your working copy of pysparse with "cvs >> diff -c -r branch-numpy" to check the changes. >> >> - Is it worth bumping the version number of pysparse to 1.0, >> since we will be breaking the name space (interface change)? >> >> - If you are compiling against numpy, the numpy header files >> don't seem to be going into the right place at least on my two >> computers. >> I haven't resolved this problem yet. For now, I was just >> copying the .h files from .../site-packages/python/numpy/core/ >> include/ to Include/Lib/numpy >> in my working copy of pysparse. >> >> Merry Christmas! >> >> -- >> Daniel Wheeler >> >> >> --------------------------------------------------------------------- >> ---- >> Take Surveys. Earn Cash. Influence the Future of IT >> Join SourceForge.net's Techsay panel and you'll get the chance to >> share your >> opinions on IT & business topics through brief surveys - and earn >> cash >> http://www.techsay.com/default.php? >> page=join.php&p=sourceforge&CID=DEVDEV_______________________________ >> ________________ >> Pysparse-users mailing list >> Pys...@li... >> https://lists.sourceforge.net/lists/listinfo/pysparse-users > > -- > Daniel Wheeler > > > ---------------------------------------------------------------------- > --- > Take Surveys. Earn Cash. Influence the Future of IT > Join SourceForge.net's Techsay panel and you'll get the chance to > share your > opinions on IT & business topics through brief surveys - and earn cash > http://www.techsay.com/default.php? > page=join.php&p=sourceforge&CID=DEVDEV________________________________ > _______________ > Pysparse-users mailing list > Pys...@li... > https://lists.sourceforge.net/lists/listinfo/pysparse-users -- Daniel Wheeler |
|
From: Daniel W. <dan...@ni...> - 2007-01-02 20:29:46
|
The changes for compatibility with numpy have now been merged back onto HEAD. I will do a new release in the near future. Cheers On Dec 21, 2006, at 12:36 PM, Daniel Wheeler wrote: > Numeric is dying slowly and being replaced by numpy. > With this in mind, I have updated pysparse to use the new numpy, > this is all on a branch at the moment. > It was straightforward matter of changing the imported header files > from Numeric's to numpy's. I am sending this email > to check to see if everybody is happy with the changes that have > been made. If there are no objections, I can merge back > onto the trunk and maybe make a new release. Changes and issues > are as follows: > > - Created a C macro defined from setup.py that either imports > numpy's or Numeric's headers depending > on the flag "use_numpy". The C macro was confusing me for > awhile, but I think it works > now. > > - I have changed the name space of itsolvers, spmatrix, jdsym > etc to all lie under pysparse. So, instead of > > >>> import spmatrix > > you do > > >>> from pysparse import spmatrix > > This change stops site-packages from being polluted. At the > moment the branch version is using pysparse_numpy > or pysparse_Numeric as the pysparse name space for testing > purposes. > > - The branch can be checked out via: > > $ cvs -d:ext:ui...@py...:/cvsroot/pysparse > co -r branch-numpy pysparse > > - I have updated the Examples and Test directories to import > from pysparse. They all seem to work apart from stuff that was > broken already. > > - I got rid of pysparse_version.py and replaced it with an > __init__.py in Lib, which does the imports and holds a __version__ > variable. > So you can now do: > > >>> import pysparse_numpy > >>> print pysparse_numpy.__version__ > 0.34.032-numpy > > - I think the default flag should be use_numpy=True since most > users will now be getting numpy rather than Numeric. This could > be automated with a try and except, which tests for the > availability of numpy or Numeric. What do you think? > > - You can diff from your working copy of pysparse with "cvs > diff -c -r branch-numpy" to check the changes. > > - Is it worth bumping the version number of pysparse to 1.0, > since we will be breaking the name space (interface change)? > > - If you are compiling against numpy, the numpy header files > don't seem to be going into the right place at least on my two > computers. > I haven't resolved this problem yet. For now, I was just > copying the .h files from .../site-packages/python/numpy/core/ > include/ to Include/Lib/numpy > in my working copy of pysparse. > > Merry Christmas! > > -- > Daniel Wheeler > > > ---------------------------------------------------------------------- > --- > Take Surveys. Earn Cash. Influence the Future of IT > Join SourceForge.net's Techsay panel and you'll get the chance to > share your > opinions on IT & business topics through brief surveys - and earn cash > http://www.techsay.com/default.php? > page=join.php&p=sourceforge&CID=DEVDEV________________________________ > _______________ > Pysparse-users mailing list > Pys...@li... > https://lists.sourceforge.net/lists/listinfo/pysparse-users -- Daniel Wheeler |
|
From: Daniel W. <dan...@ni...> - 2006-12-21 17:39:09
|
Numeric is dying slowly and being replaced by numpy.
With this in mind, I have updated pysparse to use the new numpy, this
is all on a branch at the moment.
It was straightforward matter of changing the imported header files
from Numeric's to numpy's. I am sending this email
to check to see if everybody is happy with the changes that have been
made. If there are no objections, I can merge back
onto the trunk and maybe make a new release. Changes and issues are
as follows:
- Created a C macro defined from setup.py that either imports
numpy's or Numeric's headers depending
on the flag "use_numpy". The C macro was confusing me for
awhile, but I think it works
now.
- I have changed the name space of itsolvers, spmatrix, jdsym etc
to all lie under pysparse. So, instead of
>>> import spmatrix
you do
>>> from pysparse import spmatrix
This change stops site-packages from being polluted. At the
moment the branch version is using pysparse_numpy
or pysparse_Numeric as the pysparse name space for testing
purposes.
- The branch can be checked out via:
$ cvs -d:ext:ui...@py...:/cvsroot/pysparse
co -r branch-numpy pysparse
- I have updated the Examples and Test directories to import
from pysparse. They all seem to work apart from stuff that was broken
already.
- I got rid of pysparse_version.py and replaced it with an
__init__.py in Lib, which does the imports and holds a __version__
variable.
So you can now do:
>>> import pysparse_numpy
>>> print pysparse_numpy.__version__
0.34.032-numpy
- I think the default flag should be use_numpy=True since most
users will now be getting numpy rather than Numeric. This could
be automated with a try and except, which tests for the
availability of numpy or Numeric. What do you think?
- You can diff from your working copy of pysparse with "cvs diff
-c -r branch-numpy" to check the changes.
- Is it worth bumping the version number of pysparse to 1.0,
since we will be breaking the name space (interface change)?
- If you are compiling against numpy, the numpy header files
don't seem to be going into the right place at least on my two
computers.
I haven't resolved this problem yet. For now, I was just
copying the .h files from .../site-packages/python/numpy/core/
include/ to Include/Lib/numpy
in my working copy of pysparse.
Merry Christmas!
--
Daniel Wheeler
|
|
From: Jonathan G. <gu...@ni...> - 2006-08-04 12:11:41
|
On Aug 4, 2006, at 1:18 AM, David Xu wrote: > I was wondering if Pysparse will run with Numpy 1.0b1. Has anyone > tested it out yet? PySparse does not work with NumPy. > I know Pysparse works well with Numeric-24, but I > was brave enough to try install Numpy with Pysparse fearing this could > mess up my already stable Pyspase system. There is a convertion module > to handle Numeric -> Numpy, The converter only applies to changing the syntax of Python scripts. It doesn't help convert any of the C code over to the new array type. |
|
From: David X. <dx...@my...> - 2006-08-04 05:18:12
|
Hi, I was wondering if Pysparse will run with Numpy 1.0b1. Has anyone tested it out yet? I know Pysparse works well with Numeric-24, but I was brave enough to try install Numpy with Pysparse fearing this could mess up my already stable Pyspase system. There is a convertion module to handle Numeric -> Numpy, http://projects.scipy.org/scipy/numpy/browser/trunk/numpy/lib/convertcode.py I don't how well it works and if it does work, would it bring in performance overhead? Thanks, David |
|
From: David X. <dx...@my...> - 2006-07-24 01:20:54
|
Hi, I noticed that if preconditioning (jacobi or ssor) is used along with jdsym, the eigenvectors results (Q) are off (or incorrect) compared to those calculated without preconditioning. Is that normal? How can I obtain correct results with preconditioning? Thanks, David |
|
From: Roman G. <rom...@gm...> - 2006-03-07 09:27:42
|
Hello Back when pysparse's docs where written, "Numeric" and "NumPy" referred to the same extension package, which is now called only "Numeric". The new package now called "NumPy" is likely not compatible to pysparse due to changes in the C API. I hope that upgrading to latest Numeric solves your problems. -- Roman 2006/3/6, Daniel Wheeler <dan...@ni...>: > > > Begin forwarded message: > > From: Daniel Wheeler <dan...@ni...> > Date: March 6, 2006 11:29:52 AM EST > To: Dong Xu <do...@sc...> > Subject: Re: [Pysparse-users] Pysparse compilation question > > My suggestion would be to install the latest version of Numeric (version > 24.2). > > > > http://sourceforge.net/project/showfiles.php?group_id=3D1369&package_id= =3D1351 > > > Not numpy and not numarray but Numeric. Judging by the version number you > actually installed > Numeric version 20.3 and not Numpy. Numpy does not go up to version 20.3. > But I would > still recommend the latest version of Numeric, although it may not > necessarily be causing the underlying > problem. > > > > On Mar 5, 2006, at 8:24 PM, Dong Xu wrote: > > Hi, > > I'm new user of Pysparse because i've heard about the excellent JD > eigenvalue solver in this package. I'm trying to install Pysparse on a > Ubuntu linux box. The compilation of Numpy 20.3 and Pysparse finished > with no error. I ran the Numpy test code in its demo directory, no > problem. But when I tried to run the Pysparse examples, I got the > following errors: > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D > For pysparse_test.py: > > dxu@kubuntu:~/pysparse-0.34.032/Examples$ python pysparse_test.py > ll_mat(general, [5,5]): > -------- -------- -------- -------- -------- > -------- -------- -------- -------- -------- > -------- -------- -------- -------- -------- > -------- -------- -------- -------- -------- > -------- -------- -------- -------- -------- > > 0.0 > ll_mat(general, [5,5]): > -------- -------- -------- -------- -------- > -------- -------- -------- -------- -------- > -------- -------- -------- -------- -------- > -------- -------- -------- -------- -------- > -------- -------- -------- -------- -------- > > ll_mat(general, [5,5]): > -------- -------- -------- -------- -------- > -------- -------- -------- 2.000000 -------- > -------- 1.000000 -------- -------- -------- > -------- -------- -------- -------- -------- > -------- -------- -------- -------- -------- > > csr_mat([5,5], [(1,3): 2, (2,1): 1]) > 2.0 > 0.0 > 2 > csr_mat([10,10], [(0,0): 1, (1,1): 2, (2,2): 3, (3,3): 4, (4,4): 5, > (5,5): 6, (6,6): 7, (7,7): 8, (8,8): 9, (9,9): 10]) [ 1. 1. 1. 1. > 1. 1. 1. 1. 1. 1.] [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.] > Traceback (most recent call last): > File "pysparse_test.py", line 31, in ? > A.matvec(x, y) > ValueError: arg 1 must be a 1-dimensional double array of appropriate siz= e. > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D > > For jdsym_test.py: > > dxu@kubuntu:~/pysparse-0.34.032/Examples$ python jdsym_test.py > Test 1 > JDSYM Solving A*x =3D lambda*x without preconditioning. > > N=3D 1000 ITMAX=3D 150 > KMAX=3D 5 JMIN=3D 5 JMAX=3D 10 V0DIM=3D 0 > BLKSIZE=3D 1 BLKWISE=3D FALSE > TAU=3D 0.0000e+00 JDTOL=3D 1.0000e-06 STRATEGY=3D 0 > OPTYPE=3D SYM > LINITMAX=3D 200 EPS_TR=3D 1.000e-04 TOLDECAY=3D 1.50e+00 > > Traceback (most recent call last): > File "jdsym_test.py", line 58, in ? > jmin=3D5, jmax=3D10, eps_tr=3D1e-4, clvl=3D1) > ValueError: arg 1 must be a 1-dimensional double array of appropriate siz= e. > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D > > I just couldn't figure out what went wrong with the programs. Would > you please give me some hint? > > Thank you, > > David > > > ------------------------------------------------------- > This SF.Net email is sponsored by xPML, a groundbreaking scripting langua= ge > that extends applications into web and mobile media. Attend the live webc= ast > and join the prime developer group breaking into this new coding territor= y! > http://sel.as-us.falkag.net/sel?cmd=3Dlnk&kid=110944&bid$1720&dat=121642 > _______________________________________________ > Pysparse-users mailing list > Pys...@li... > https://lists.sourceforge.net/lists/listinfo/pysparse-users > > > > ------------------------------------- > Daniel Wheeler > Telephone: (301) 975-8358 > > > > ------------------------------------- > > Daniel Wheeler > > Telephone: (301) 975-8358 > |
|
From: Daniel W. <dan...@ni...> - 2006-03-06 16:32:06
|
Begin forwarded message: > From: Daniel Wheeler <dan...@ni...> > Date: March 6, 2006 11:29:52 AM EST > To: Dong Xu <do...@sc...> > Subject: Re: [Pysparse-users] Pysparse compilation question > > My suggestion would be to install the latest version of Numeric =20 > (version 24.2). > > > http://sourceforge.net/project/showfiles.php?=20 > group_id=3D1369&package_id=3D1351 > > > Not numpy and not numarray but Numeric. Judging by the version =20 > number you actually installed > Numeric version 20.3 and not Numpy. Numpy does not go up to version =20= > 20.3. But I would > still recommend the latest version of Numeric, although it may not =20 > necessarily be causing the underlying > problem. > > On Mar 5, 2006, at 8:24 PM, Dong Xu wrote: > >> Hi, >> >> I'm new user of Pysparse because i've heard about the excellent JD >> eigenvalue solver in this package. I'm trying to install Pysparse =20 >> on a >> Ubuntu linux box. The compilation of Numpy 20.3 and Pysparse finished >> with no error. I ran the Numpy test code in its demo directory, no >> problem. But when I tried to run the Pysparse examples, I got the >> following errors: >> =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D >> For pysparse_test.py: >> >> dxu@kubuntu:~/pysparse-0.34.032/Examples$ python pysparse_test.py >> ll_mat(general, [5,5]): >> -------- -------- -------- -------- -------- >> -------- -------- -------- -------- -------- >> -------- -------- -------- -------- -------- >> -------- -------- -------- -------- -------- >> -------- -------- -------- -------- -------- >> >> 0.0 >> ll_mat(general, [5,5]): >> -------- -------- -------- -------- -------- >> -------- -------- -------- -------- -------- >> -------- -------- -------- -------- -------- >> -------- -------- -------- -------- -------- >> -------- -------- -------- -------- -------- >> >> ll_mat(general, [5,5]): >> -------- -------- -------- -------- -------- >> -------- -------- -------- 2.000000 -------- >> -------- 1.000000 -------- -------- -------- >> -------- -------- -------- -------- -------- >> -------- -------- -------- -------- -------- >> >> csr_mat([5,5], [(1,3): 2, (2,1): 1]) >> 2.0 >> 0.0 >> 2 >> csr_mat([10,10], [(0,0): 1, (1,1): 2, (2,2): 3, (3,3): 4, (4,4): 5, >> (5,5): 6, (6,6): 7, (7,7): 8, (8,8): 9, (9,9): 10]) [ 1. 1. 1. 1. >> 1. 1. 1. 1. 1. 1.] [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.] >> Traceback (most recent call last): >> File "pysparse_test.py", line 31, in ? >> A.matvec(x, y) >> ValueError: arg 1 must be a 1-dimensional double array of =20 >> appropriate size. >> =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D >> >> For jdsym_test.py: >> >> dxu@kubuntu:~/pysparse-0.34.032/Examples$ python jdsym_test.py >> Test 1 >> JDSYM Solving A*x =3D lambda*x without preconditioning. >> >> N=3D 1000 ITMAX=3D 150 >> KMAX=3D 5 JMIN=3D 5 JMAX=3D 10 V0DIM=3D 0 >> BLKSIZE=3D 1 BLKWISE=3D FALSE >> TAU=3D 0.0000e+00 JDTOL=3D 1.0000e-06 STRATEGY=3D 0 >> OPTYPE=3D SYM >> LINITMAX=3D 200 EPS_TR=3D 1.000e-04 TOLDECAY=3D 1.50e+00 >> >> Traceback (most recent call last): >> File "jdsym_test.py", line 58, in ? >> jmin=3D5, jmax=3D10, eps_tr=3D1e-4, clvl=3D1) >> ValueError: arg 1 must be a 1-dimensional double array of =20 >> appropriate size. >> =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D >> >> I just couldn't figure out what went wrong with the programs. Would >> you please give me some hint? >> >> Thank you, >> >> David >> >> >> ------------------------------------------------------- >> This SF.Net email is sponsored by xPML, a groundbreaking scripting =20= >> language >> that extends applications into web and mobile media. Attend the =20 >> live webcast >> and join the prime developer group breaking into this new coding =20 >> territory! >> http://sel.as-us.falkag.net/sel?cmd=3Dlnk&kid=110944&bid$1720&dat=12164= 2 >> _______________________________________________ >> Pysparse-users mailing list >> Pys...@li... >> https://lists.sourceforge.net/lists/listinfo/pysparse-users >> > > ------------------------------------- > Daniel Wheeler > Telephone: (301) 975-8358 > ------------------------------------- Daniel Wheeler Telephone: (301) 975-8358 |
|
From: Dong X. <do...@sc...> - 2006-03-06 01:24:43
|
Hi,
I'm new user of Pysparse because i've heard about the excellent JD
eigenvalue solver in this package. I'm trying to install Pysparse on a
Ubuntu linux box. The compilation of Numpy 20.3 and Pysparse finished
with no error. I ran the Numpy test code in its demo directory, no
problem. But when I tried to run the Pysparse examples, I got the
following errors:
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D
For pysparse_test.py:
dxu@kubuntu:~/pysparse-0.34.032/Examples$ python pysparse_test.py
ll_mat(general, [5,5]):
-------- -------- -------- -------- --------
-------- -------- -------- -------- --------
-------- -------- -------- -------- --------
-------- -------- -------- -------- --------
-------- -------- -------- -------- --------
0.0
ll_mat(general, [5,5]):
-------- -------- -------- -------- --------
-------- -------- -------- -------- --------
-------- -------- -------- -------- --------
-------- -------- -------- -------- --------
-------- -------- -------- -------- --------
ll_mat(general, [5,5]):
-------- -------- -------- -------- --------
-------- -------- -------- 2.000000 --------
-------- 1.000000 -------- -------- --------
-------- -------- -------- -------- --------
-------- -------- -------- -------- --------
csr_mat([5,5], [(1,3): 2, (2,1): 1])
2.0
0.0
2
csr_mat([10,10], [(0,0): 1, (1,1): 2, (2,2): 3, (3,3): 4, (4,4): 5,
(5,5): 6, (6,6): 7, (7,7): 8, (8,8): 9, (9,9): 10]) [ 1. 1. 1. 1.
1. 1. 1. 1. 1. 1.] [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]
Traceback (most recent call last):
File "pysparse_test.py", line 31, in ?
A.matvec(x, y)
ValueError: arg 1 must be a 1-dimensional double array of appropriate size.
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D
For jdsym_test.py:
dxu@kubuntu:~/pysparse-0.34.032/Examples$ python jdsym_test.py
Test 1
JDSYM Solving A*x =3D lambda*x without preconditioning.
N=3D 1000 ITMAX=3D 150
KMAX=3D 5 JMIN=3D 5 JMAX=3D 10 V0DIM=3D 0
BLKSIZE=3D 1 BLKWISE=3D FALSE
TAU=3D 0.0000e+00 JDTOL=3D 1.0000e-06 STRATEGY=3D 0
OPTYPE=3D SYM
LINITMAX=3D 200 EPS_TR=3D 1.000e-04 TOLDECAY=3D 1.50e+00
Traceback (most recent call last):
File "jdsym_test.py", line 58, in ?
jmin=3D5, jmax=3D10, eps_tr=3D1e-4, clvl=3D1)
ValueError: arg 1 must be a 1-dimensional double array of appropriate size.
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D
I just couldn't figure out what went wrong with the programs. Would
you please give me some hint?
Thank you,
David
|
|
From: Daniel W. <dan...@ni...> - 2005-09-06 14:36:22
|
I have been having some problems building and using pysparse with
Fedora Core 4.
Fedora Core 4 doesn't seem to have libg2c.so. Reading around and
talking to people
I got the impression that blas and lapack have implicilty included
g2c in the versions for Fedora Core 4.
Thus, I removed g2c from the libraries list in setup.py. Build and
install seems to work. However, when I import modules I get the
following error.
>>> import precon
ImportError: /usr/lib/libblas.so.2: undefined symbol: e_wsfe
I'm not sure what to do next maybe create a link from /usr/lib/
libg2c.so.0.0.0 to /usr/libg2c.so
and re-include g2c. Does anyone have experience with this issue?
Thanks
-------------------------------------
Daniel Wheeler
Telephone: (301) 975-8358
|
|
From: Roman G. <rom...@gm...> - 2005-06-22 12:10:52
|
Hi Implementing a C API for a Python extension is one of the more obscure programming tasks IMHO. To expose additional pysparse functions using the C API the spmatrix_api.h needs to be extended. This is however an tedious and error prone task. I wrote a simple Python script to generate spmatrix_api.h. You can play around with this script if you want. (see attached file) Cheers -- Roman On 6/22/05, Dominique Orban <dom...@gm...> wrote: > On 6/22/05, Roman Geus <rom...@gm...> wrote: > > Hi Dominique > > > > Sorry for not replying. I actually forgot to register myself for this l= ist > > :-/ > > > > > Hello, > > > > > > I am linking my C code to PySparse and would like to create a CSR > > > matrix using newCSRMatObject. I don"t see why the prototype for that > > > function does not appear in the headers unless SPMATRIX_MODULE is > > > #define"d. > > > > > If I #define SPMATRIX_MODULE before #include-ing csr_mat.h, I obtain > > > tons of warning and error messages. > > > > newCSRMatObject() is not supposed to be visible outside of pysparse. > > > > > Why not have newCSRMatObject"s prototype in spmatrix_api.h like, e.g.= , > > > newLLMatObject? Same for newSSSMatObject. > > > > Since there is no way to create CSR and SSS matrices, other than > > building an LL matrix first and then converting it, I don't really see > > any benefit in exposing these functions. Am I missing something? >=20 > Hello Roman, >=20 > There may not be any benefits from Python. However, from C, you could > obviously create instances of CSR and SSS matrices without the > overhead of first creating an ll_mat and then converting it (which > takes some time when the matrices get large). If you know in advance > that you're going to create a matrix that will not be modified (in my > case: the factors of a system) but that you will use (eg for > backsolves), it would be nice to be able to skip the overhead of > creating an ll_mat. >=20 > I see it as a limitation of PySparse right now because my C kernel > isn't able to call directly other low-level language subroutines that > expect CSR or SSS matrices. Instead, I must take the long route > through ll_mat land and that has time and computational consequences. >=20 > I am not sure how to expose those functions though as tweaking the > source code produces lots of errors and warnings. I must not be > tweaking right. >=20 > Dominique > |
|
From: Dominique O. <dom...@gm...> - 2005-06-22 11:43:49
|
On 6/22/05, Roman Geus <rom...@gm...> wrote: > Hi Dominique >=20 > Sorry for not replying. I actually forgot to register myself for this lis= t > :-/ >=20 > > Hello, > >=20 > > I am linking my C code to PySparse and would like to create a CSR > > matrix using newCSRMatObject. I don"t see why the prototype for that > > function does not appear in the headers unless SPMATRIX_MODULE is > > #define"d. > =20 > > If I #define SPMATRIX_MODULE before #include-ing csr_mat.h, I obtain > > tons of warning and error messages. >=20 > newCSRMatObject() is not supposed to be visible outside of pysparse. > =20 > > Why not have newCSRMatObject"s prototype in spmatrix_api.h like, e.g., > > newLLMatObject? Same for newSSSMatObject. >=20 > Since there is no way to create CSR and SSS matrices, other than > building an LL matrix first and then converting it, I don't really see > any benefit in exposing these functions. Am I missing something? Hello Roman, There may not be any benefits from Python. However, from C, you could obviously create instances of CSR and SSS matrices without the overhead of first creating an ll_mat and then converting it (which takes some time when the matrices get large). If you know in advance that you're going to create a matrix that will not be modified (in my case: the factors of a system) but that you will use (eg for backsolves), it would be nice to be able to skip the overhead of creating an ll_mat. I see it as a limitation of PySparse right now because my C kernel isn't able to call directly other low-level language subroutines that expect CSR or SSS matrices. Instead, I must take the long route through ll_mat land and that has time and computational consequences. I am not sure how to expose those functions though as tweaking the source code produces lots of errors and warnings. I must not be tweaking right. Dominique |
1 message has been excluded from this view by a project administrator.
