Date:         Thu, 20 Jan 2005 11:29:21 -0300
Reply-To:     Hector Maletta <hmaletta@fibertel.com.ar>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Hector Maletta <hmaletta@fibertel.com.ar>
Subject:      Re: Vector multiplication: a clarification
Comments: To: "Peck, Jon" <peck@spss.com>
In-Reply-To:  <5CFEFDB5226CB54CBB4328B9563A12EEDB5EAB@hqemail2.spss.com>
Content-Type: text/plain; charset="US-ASCII"

Jon,
Thanks for the tips. In my case I would do a vectorial multiplication of one
vector of variables for another vector of variables, FOR EACH CASE. It is
essentially a COMPUTE transformation, namely the sum of a series of
products. Therefore I imagine the size of the dataset (i.e. the number of
cases) is not necessarily an issue.
Can I do what I want without using a LOOP structure, just multiplying the
two vectors of variables, say X and W, to obtain a third set of variables,
say Z? What I imagine is something like this:
VECTOR Z = Z1 TO Z50.
VECTOR X = X1 TO X50.
VECTOR W = W1 TO W50.
COMPUTE Z = X * W.

Hector

> -----Original Message-----
> From: SPSSX(r) Discussion [mailto:SPSSX-L@LISTSERV.UGA.EDU]
> On Behalf Of Peck, Jon
> Sent: Thursday, January 20, 2005 11:18 AM
> To: SPSSX-L@LISTSERV.UGA.EDU
> Subject: Re: Vector multiplication: a clarification
>
>
> The matrix language contains all the usual matrix operations
> as well as elementwise operators and many more complex
> functions such as Cholesky decomposition, determinants,
> eigenvalues/vectors, and SVD, just to name a few.  And it has
> programming structures such as loops.
>
> You should be aware, though, that Matrix, unlike most SPSS
> procedures keeps its data in memory, so very large datasets
> will require lots of memory (or big swap files and patience).
>  But lots of memory is pretty common these days.
>
> Regards,
> Jon Peck
>
> -----Original Message-----
> From: SPSSX(r) Discussion [mailto:SPSSX-L@LISTSERV.UGA.EDU]
> On Behalf Of Mark A Davenport MADAVENP
> Sent: Thursday, January 20, 2005 7:36 AM
> To: SPSSX-L@LISTSERV.UGA.EDU
> Subject: Re: [SPSSX-L] Vector multiplication: a clarification
>
> Oops, I should read messages more closely.  It does appear
> that SPSS matrix notation will handle vectors and scalars.  I
> have only used it once with 2 matrices. SPSSX-L@VM.MARIST.EDU cc
>
> Subject
> Re: Vector multiplication
>
>
>
>
>
>
> About my question on multiplying two vectors of variables,
> Richard Ristow
> wrote:
> > Unless I'm much mistaken, there's no way to manipulate a
> vector as a
> > unit.
>
> I supposed the matrix language available in SPSS, which I am
> not very much acquainted with, may have some way of doing
> vector and matrix algebra operations, such as vector
> multiplication. It does apparently not. Am I wrong?
>
> Hector
>