numpy-discussion — Discussion list for all users of Numerical Python

Re: [Numpy-discussion] How do I make a diagonal matrix?

[Joris De R. <jo...@st...>]

On Friday 23 June 2006 16:34, Sven Schreiber wrote:
   [SS]: Keith Goodman schrieb:
   [SS]: > How do I make a NxN diagonal matrix with a Nx1 column vector x along
   [SS]: > the diagonal?
   [SS]: > 
   [SS]: 
   [SS]: >>> help(n.diag)
   [SS]: Help on function diag in module numpy.lib.twodim_base:
   [SS]: 
   [SS]: diag(v, k=0)
   [SS]:     returns the k-th diagonal if v is a array or returns a array
   [SS]:     with v as the k-th diagonal if v is a vector.

See also the Numpy Example List for a few examples:
http://www.scipy.org/Numpy_Example_List

J.

Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm

Re: [Numpy-discussion] How do I make a diagonal matrix?

[Sven S. <sve...@gm...>]

Keith Goodman schrieb:
> How do I make a NxN diagonal matrix with a Nx1 column vector x along
> the diagonal?
> 

>>> help(n.diag)
Help on function diag in module numpy.lib.twodim_base:

diag(v, k=0)
    returns the k-th diagonal if v is a array or returns a array
    with v as the k-th diagonal if v is a vector.

[Numpy-discussion] How do I make a diagonal matrix?

[Keith G. <kwg...@gm...>]

How do I make a NxN diagonal matrix with a Nx1 column vector x along
the diagonal?

Re: [Numpy-discussion] fromfile croaking on windows

[Albert S. <fu...@gm...>]

Hello all

Travis Oliphant wrote:
> Mathew Yeates wrote:
> > when I try and load a file with numpy.fromfile I keep getting a message
> ....
> > 7245092 items requested but only 3899 read. Its always the same number
> read.
> >
> >
> Which platform are you on?   Could you show exactly how you are calling
> the function.
> 
> There were some reports of strange behavior on Windows that may be
> related to file-locking.  I'm just not sure at this point.

I did some experiments.

With my test file, this always fails:

y = N.fromfile('temp.dat', dtype=N.float64)

This works:

y = N.fromfile(file('temp.dat','rb'), dtype=N.float64)

More details in this ticket:

http://projects.scipy.org/scipy/numpy/ticket/103

I don't quite understand how file-locking can be causing these problems.
Travis, care to elaborate on what you think might be causing these problems?

Cheers,

Albert

[Numpy-discussion] eye and identity: why both?

[Sven S. <sve...@gm...>]

identity seems to be a "crippled" version of eye without any value
added, apart from backwards-compatibility;

So from a user point of view, which one does numpy recommend?

And from a developer point of view (which doesn't really apply to me, of
course), should identity maybe become an alias for eye(n, dtype=...)?

Or is there a subtle (or not so subtle...) difference I am missing?

I am aware this question is not really that important since everything
works, but when I read that there will be a 1.0beta soon I thought maybe
this is the right time to ask those kind of questions.

Here are the help-strings:

eye(N, M=None, k=0, dtype=<type 'int32scalar'>)
    eye returns a N-by-M 2-d array where the  k-th diagonal is all ones,
    and everything else is zeros.

identity(n, dtype=<type 'int32scalar'>)
    identity(n) returns the identity 2-d array of shape n x n.

Cheers,
Sven

Re: [Numpy-discussion] fromfile croaking on windows

[Travis O. <oli...@ie...>]

Mathew Yeates wrote:
> when I try and load a file with numpy.fromfile I keep getting a message ....
> 7245092 items requested but only 3899 read. Its always the same number read.
>
>   
Which platform are you on?   Could you show exactly how you are calling 
the function. 

There were some reports of strange behavior on Windows that may be 
related to file-locking.  I'm just not sure at this point.

-Travis

[Numpy-discussion] p.s. Re: fromfile croaking on windows

[Mathew Y. <my...@jp...>]

When I specify count=-1 I get the exact same error. So, numpy was able 
to determine the filesize. It just can't read it.

Mathew

Mathew Yeates wrote:
> when I try and load a file with numpy.fromfile I keep getting a message ....
> 7245092 items requested but only 3899 read. Its always the same number read.
>
> I've checked and I'm giving the correct filename and its the correct 
> size. Any idea whats going on?
> This is with 0.9.8
>
> Mathew
>
>
>
> Using Tomcat but need to do more? Need to support web services, security?
> Get stuff done quickly with pre-integrated technology to make your job easier
> Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
> _______________________________________________
> Numpy-discussion mailing list
> Num...@li...
> https://lists.sourceforge.net/lists/listinfo/numpy-discussion
>
>   

[Numpy-discussion] fromfile croaking on windows

[Mathew Y. <my...@jp...>]

when I try and load a file with numpy.fromfile I keep getting a message ....
7245092 items requested but only 3899 read. Its always the same number read.

I've checked and I'm giving the correct filename and its the correct 
size. Any idea whats going on?
This is with 0.9.8

Mathew

Re: [Numpy-discussion] Recent SVN of NumPy has issues with SciPy

[Travis O. <oli...@ie...>]

Travis Oliphant wrote:
> There are still some issues with my recent check-in for NumPy (r2663).   
> But, it does build and run the numpy.tests cleanly.  (It's failing on 
> SciPy tests...)
>   
These issues are now fixed (it was a brain-dead optimization that just 
doesn't work and was only exposed when converting between C- and 
Fortran- arrays during a cast..

Feel free to use SVN again...  I do like to keep SVN so that it works.

-Travis

Re: [Numpy-discussion] MA bug or feature?

[Michael S. <mic...@gm...>]

On 6/23/06, Pierre GM <pgm...@ma...> wrote:
> On Wednesday 21 June 2006 22:01, Michael Sorich wrote:
> > Nevertheless, the differences between a masked array with a boolean
> > mask and a mask of booleans have caused me trouble before. Especially
> > when there are hidden in-place conversions of a mask which is a array
> > of False to a mask which is False. e.g.
>
> OK, I'm still using 0.9.8 and I can't help you with this one. In that version,
> N.asarray transforms the MA into a ndarray, so you lose the mask.

No, the mask of ma1 is converted in place to False. ma1 remains a MaskedArray

import numpy
ma1 = numpy.ma.array(((1.,2,3),(4,5,6)), mask=((0,0,0),(0,0,0)))
print ma1.mask, type(ma1)
numpy.asarray(ma1)
print ma1.mask, type(ma1)

--output--
[[False False False]
 [False False False]] <class 'numpy.core.ma.MaskedArray'>
False <class 'numpy.core.ma.MaskedArray'>

> But I wonder: if none of your values are masked, the natural behavior would be
> to have `data.mask==nomask`, which speeds up things a bit. This gain of time
> is why I was suggesting that `mask` would be forced to `nomask` at the
> creation, if `mask.any()==False`.
>
> Could you give me some examples of cases where you need the mask to stay as an
> array of False ?
> If you need to access the mask as an array, you can always use
> MA.getmaskarray.

If it did not sometimes effect the behaviour of the masked array, I
would not be worried about automatic conversions between the two forms
of the mask. Is it agreed that there should not be any differences in
the behavior of the two forms of masked array e.g. with a mask of
[[False,False],[False,False]] vs False? It is frustrating to track
down exceptions when the array has one behavior, then there is a
implicit conversion of the mask which changes the behaviour of the
array.

Mike

Re: [Numpy-discussion] problem building Numeric on python 2.5

[Tim H. <tim...@co...>]

Russell E. Owen wrote:
> I just installed python 2.5b1 on my Mac (10.4 ppc) and can't seem to get 
> Numeric 24.2 installed. It seems to build fine (no obvious error 
> messages), but when I try to import it I get:
> Python 2.5b1 (r25b1:47038M, Jun 20 2006, 16:17:55) 
> [GCC 4.0.1 (Apple Computer, Inc. build 5341)] on darwin
> Type "help", "copyright", "credits" or "license" for more information.
>   
>>>> import Numeric
>>>>         
> __main__:1: ImportWarning: Not importing directory 
> '/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-pac
> kages/Numeric': missing __init__.py
>   
>
> Any ideas? Is it somehow incompatible with python 2.5b1?
>   
Import warning is a new 'feature' of 2.5. It warns if there are 
directories on sys.path that are *not* packages. I'll refer you to the 
py-dev archives if you want figure out the motivation for that. So, if 
everything seems to work, there's a good chance that nothing's wrong, 
but that your just seeing a complaint due to this new behaviour. If you 
check recent messages on Python-dev someone just posted a recipe for 
suppressing this warning.

-tim


> For what it's worth, numarray builds and installs fine. I've not tried 
> numpy or any other packages yet.
>
> -- Russell
>
>
> Using Tomcat but need to do more? Need to support web services, security?
> Get stuff done quickly with pre-integrated technology to make your job easier
> Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
> _______________________________________________
> Numpy-discussion mailing list
> Num...@li...
> https://lists.sourceforge.net/lists/listinfo/numpy-discussion
>
>
>   

Re: [Numpy-discussion] problem building Numeric on python 2.5

[Robert K. <rob...@gm...>]

Russell E. Owen wrote:
> I just installed python 2.5b1 on my Mac (10.4 ppc) and can't seem to get 
> Numeric 24.2 installed. It seems to build fine (no obvious error 
> messages), but when I try to import it I get:
> Python 2.5b1 (r25b1:47038M, Jun 20 2006, 16:17:55) 
> [GCC 4.0.1 (Apple Computer, Inc. build 5341)] on darwin
> Type "help", "copyright", "credits" or "license" for more information.
>>>> import Numeric
> __main__:1: ImportWarning: Not importing directory 
> '/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-pac
> kages/Numeric': missing __init__.py
> 
> Any ideas? Is it somehow incompatible with python 2.5b1?
> 
> For what it's worth, numarray builds and installs fine. I've not tried 
> numpy or any other packages yet.

See Konrad Hinsen's post earlier today "Numeric and Python 2.5b1" for a 
description of the issue and a way to silence the warnings. It's just a warning, 
though, not an error.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless enigma
  that is made terrible by our own mad attempt to interpret it as though it had
  an underlying truth."
   -- Umberto Eco

[Numpy-discussion] problem building Numeric on python 2.5

[Russell E. O. <ro...@ce...>]

I just installed python 2.5b1 on my Mac (10.4 ppc) and can't seem to get 
Numeric 24.2 installed. It seems to build fine (no obvious error 
messages), but when I try to import it I get:
Python 2.5b1 (r25b1:47038M, Jun 20 2006, 16:17:55) 
[GCC 4.0.1 (Apple Computer, Inc. build 5341)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>> import Numeric
__main__:1: ImportWarning: Not importing directory 
'/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-pac
kages/Numeric': missing __init__.py
>>> 

Any ideas? Is it somehow incompatible with python 2.5b1?

For what it's worth, numarray builds and installs fine. I've not tried 
numpy or any other packages yet.

-- Russell

Re: [Numpy-discussion] sourceforge advertising

[Robert K. <rob...@gm...>]

Stefan van der Walt wrote:
> Hi,
> 
> I noticed that sourceforge now adds another 8 lines of advertisement
> to the bottom of every email sent to the list.  Am I the only one who
> finds this annoying?  Is there any reason why the numpy list can't run
> on scipy.org?

We'd be happy to move it to scipy.org. However moving a mailing list is always a 
hassle for subscribers, so we were not going to bother until there was a 
compelling reason. This may be one, though.

For all subscribers: If you have an opinion over whether to move the list or to 
keep it on Sourceforge, please email me *offlist*. If enough people want to move 
and few people want to stay, we'll set up a new mailing list on scipy.org.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless enigma
  that is made terrible by our own mad attempt to interpret it as though it had
  an underlying truth."
   -- Umberto Eco

[Numpy-discussion] sourceforge advertising

[Stefan v. d. W. <st...@su...>]

Hi,

I noticed that sourceforge now adds another 8 lines of advertisement
to the bottom of every email sent to the list.  Am I the only one who
finds this annoying?  Is there any reason why the numpy list can't run
on scipy.org?

Regards
St=E9fan

Re: [Numpy-discussion] MA bug or feature?

[Paul D. <pfd...@gm...>]

Pierre wrote:


> I agree that's a bit confusing here, and there might be some room for
> improvement (for example, changing the current
> `if m is nomask` to `if m is nomask or m.any()==False`, or better, forcing
> mask to nomask if mask.any()==False). But I don;t think that qualifies as
> bug.


In the original MA in Numeric, I decided that to constantly check for masks
that didn't actually mask anything was not a good idea. It punishes normal
use with a very expensive check that is rarely going to be true.

If you are in a setting where you do not want this behavior, but instead
want masks removed whenever possible, you may wish to wrap or replace things
like masked_array so that they call make_mask with flag = 1:

y = masked_array(data, make_mask(maskdata, flag=1))

y will have no mask if maskdata is all false.

Thanks to Pierre for pointing out about masked_print_option.

Paul

[Numpy-discussion] problem building NumPy

[<saa...@sf...>]

Hi

I've never done a compile of a program before so I followed the instructions
found at scipy.org/Build_for_Windows, but when I attempt to compile the
latest version I get many many lines that say "could not locate executable
ifort/ifc/efc/g77/g95" etc. 

And when I run the test "numpy.test()" numpy cannot be found, so I'm
assuming that I haven't actually installed anything.

Any advise and direction would be greatly appreciated

best regards
Sonja Aagesen
MA Candidate
Dept of Archaeology
Simon Fraser University
saa...@sf...

Re: [Numpy-discussion] How do I seed the radom number generator?

[Keith G. <kwg...@gm...>]

On 6/22/06, Robert Kern <rob...@gm...> wrote:
> Keith Goodman wrote:
> > How do I seed rand and randn?
>
> If you can, please use the .rand() and .randn() methods on a RandomState object
> which you can initialize with whatever seed you like.
>
> In [1]: import numpy as np
> rs
> In [2]: rs = np.random.RandomState([12345678, 90123456, 78901234])
>
> In [3]: rs.rand(5)
> Out[3]: array([ 0.40355172,  0.27449337,  0.56989746,  0.34767024,  0.47185004])

Perfect! Thank you.

Re: [Numpy-discussion] How do I seed the radom number generator?

[Robert K. <rob...@gm...>]

Keith Goodman wrote:
> How do I seed rand and randn?

If you can, please use the .rand() and .randn() methods on a RandomState object 
which you can initialize with whatever seed you like.

In [1]: import numpy as np
rs
In [2]: rs = np.random.RandomState([12345678, 90123456, 78901234])

In [3]: rs.rand(5)
Out[3]: array([ 0.40355172,  0.27449337,  0.56989746,  0.34767024,  0.47185004])

In [5]: np.random.RandomState.seed?
Type:           method_descriptor
Base Class:     <type 'method_descriptor'>
String Form:    <method 'seed' of 'mtrand.RandomState' objects>
Namespace:      Interactive
Docstring:
     Seed the generator.

     seed(seed=None)

     seed can be an integer, an array (or other sequence) of integers of any
     length, or None. If seed is None, then RandomState will try to read data
     from /dev/urandom (or the Windows analogue) if available or seed from
     the clock otherwise.


The rand() and randn() "functions" are actually references to methods on a 
global instance of RandomState. The .seed() method on that object is also 
similarly exposed as numpy.random.seed(). If you are writing new code, please 
explicitly use a RandomState object. Only use numpy.random.seed() if you must 
control code that uses the global rand() and randn() "functions" and you can't 
modify it.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless enigma
  that is made terrible by our own mad attempt to interpret it as though it had
  an underlying truth."
   -- Umberto Eco

Re: [Numpy-discussion] C-API support for numarray added to NumPy

[Christopher H. <ch...@st...>]

>You will also need to change the include directories used in compiling 
>by appending the directories returned by 
>numpy.numarray.util.get_numarray_include_dirs()
>

Hi Travis,

I believe that there is a problem with this function.  When executing interactively with numpy version 0.9.9.2660 I get the following result:

Python 2.4.1 (#65, Mar 30 2005, 09:13:57) [MSC v.1310 32 bit (Intel)]
Type "copyright", "credits" or "license" for more information.

In [1]: import numpy

In [2]: numpy.__version__
Out[2]: '0.9.9.2660'

In [3]: import numpy.numarray.util as nnu

In [4]: nnu.get_numarray_include_dirs()
Out[4]: ['C:\\Python24\\lib\\site-packages\\numpy\\numarray']

Unfortunately this does not have the appropriate (or any) header files.

Chris

[Numpy-discussion] How do I seed the radom number generator?

[Keith G. <kwg...@gm...>]

How do I seed rand and randn?

Re: [Numpy-discussion] MA bug or feature?

[Pierre GM <pgm...@ma...>]

On Wednesday 21 June 2006 22:01, Michael Sorich wrote:
> I was setting the fill_value as 'NA' when constructing the array so
> the masked values would be printed as 'NA'. It is not a big deal to
> avoid doing this.

You can use masked_print_option, as illustrated below, without using a 
fill_value incompatible with your data type.

>>>import numpy.core.ma as MA
>>>X = MA.array([1,2,3],maks=[0,1,0])
>>>print X
[1 -- 3]
>>>MA.masked_print_option=MA._MaskedPrintOption('N/A')
>>>print X
[1 N/A 3]


> Nevertheless, the differences between a masked array with a boolean
> mask and a mask of booleans have caused me trouble before. Especially
> when there are hidden in-place conversions of a mask which is a array
> of False to a mask which is False. e.g.

OK, I'm still using 0.9.8 and I can't help you with this one. In that version,  
N.asarray transforms the MA into a ndarray, so you lose the mask.

But I wonder: if none of your values are masked, the natural behavior would be 
to have `data.mask==nomask`, which speeds up things a bit. This gain of time 
is why I was suggesting that `mask` would be forced to `nomask` at the 
creation, if `mask.any()==False`.

Could you give me some examples of cases where you need the mask to stay as an 
array of False ?
If you need to access the mask as an array, you can always use 
MA.getmaskarray.

Re: [Numpy-discussion] Matrix construction

[Bill B. <wb...@gm...>]

On 6/23/06, Keith Goodman <kwg...@gm...> wrote:
>
> On 6/22/06, Bill Baxter <wb...@gm...> wrote:
> > On 6/22/06, Ed Schofield <sch...@ft...> wrote:
> >
> > >
> > > On 22/06/2006, at 12:40 AM, Bill Baxter wrote:
> > >
> > > > Actually I think using mat() (just an alias for the matrix
> > > > constructor) is a bad way to do it.  That mat() (and most others on
> > > > that page) should probably be replaced with asmatrix() to avoid the
> > > > copy.
> > >
> > > Perhaps the 'mat' function should become an alias for 'asmatrix'.
> > > I've thought this for a while.
> >
> >
> > That makes sense to me.  As far as I know, asmatrix() defaults to
> calling
> > the constructor if it can't snarf the memory of the object being passed
> in.
> >
> > So, go on, shoot Ed and me down! :-)
>
> I can anticipate one problem: the Pirates will want their three-letter
> abbreviation for asarray.


arr() me maties!

Will functions like rand and eye always return arrays? Or will there
> be a day when you can tell numpy that you are working with matrices
> and then it will return matrices when you call rand, eye, etc?
>

I don't disagree there's a need, but you can always make your own:

def mrand(*vargs):
    return asmatrix(rand(*vargs))

def meye(N, **kwargs):
    return asmatrix(eye(N,**kwargs))

--bb

[Numpy-discussion] Recent SVN of NumPy has issues with SciPy

[Travis O. <oli...@ie...>]

There are still some issues with my recent check-in for NumPy (r2663).   
But, it does build and run the numpy.tests cleanly.  (It's failing on 
SciPy tests...)

You may want to hold off for a few hours until I can straighten it out.

-Travis

Re: [Numpy-discussion] Matrix construction

[Keith G. <kwg...@gm...>]

On 6/22/06, Bill Baxter <wb...@gm...> wrote:
> On 6/22/06, Ed Schofield <sch...@ft...> wrote:
>
> >
> > On 22/06/2006, at 12:40 AM, Bill Baxter wrote:
> >
> > > Actually I think using mat() (just an alias for the matrix
> > > constructor) is a bad way to do it.  That mat() (and most others on
> > > that page) should probably be replaced with asmatrix() to avoid the
> > > copy.
> >
> > Perhaps the 'mat' function should become an alias for 'asmatrix'.
> > I've thought this for a while.
>
>
> That makes sense to me.  As far as I know, asmatrix() defaults to calling
> the constructor if it can't snarf the memory of the object being passed in.
>
> So, go on, shoot Ed and me down! :-)

I can anticipate one problem: the Pirates will want their three-letter
abbreviation for asarray.

Will functions like rand and eye always return arrays? Or will there
be a day when you can tell numpy that you are working with matrices
and then it will return matrices when you call rand, eye, etc?