How to group Cluster Analysis outputs

classic Classic list List threaded Threaded
3 messages Options
CJG
Reply | Threaded
Open this post in threaded view
|

How to group Cluster Analysis outputs

CJG
Hello,

I am using version 11 & have a quick question about cluster analyses. I
have a set of data (individuals & their measurements) which I would like
to group together. A cluster analysis will do that, but in the case of
large numbers of individuals &/or tight groupings, I cannot read the ID
numbers that are output to the screen (in the cluster graph). This is very
important as I would like to be able to check a map (in a different
program) to see if the resulting cluster analysis groups are
geographically grouped as well.

So, in short, how can I get SPSS to identify the individual IDs in each
cluster, or at least add a new field to each individual record with a
cluster number? Hopefuly that makes sense... Thanks,

Chris
Reply | Threaded
Open this post in threaded view
|

Re: How to group Cluster Analysis outputs

Art Kendall-2
An excellent simple clustering (as opposed to hierarchical clustering)
TWOSTEP was introduced in 11.5.
The AIC and BIC stats will help you identifying what number of cluster
solutions to examine to determine which solution to retain.
Use
  /SAVE CLUSTER [VARIABLE = varname]]

I cannot recall for sure whether there was a charge to upgrade from
11.0, but I don't think so.
If you cannot get 11.5 and TWOSTEP try a few runs of QUICK CLUSTER.

in Quick Cluster use /SAVE=CLUSTER(RUN1)
for the first random sort of your cases
  /SAVE=CLUSTER(RUN2)
etc for 4 or 5 runs of Quick cluster with different numbers of clusters and

If you have only a small number a cases to cluster and want to see
hierarchical solutions cluster use /save=cluster(3,5)
to save memberships for levels with three to 5 clusters.
see the documentation for the /METHOD for how to supply the root name
for memberships.

You can see the documentation by highlighting the word CLUSTER or
TWOSTEP, etc. in the syntax window and clicking the syntax icon.

Art Kendall
Social Research Consultants

CJG wrote:

> Hello,
>
> I am using version 11 & have a quick question about cluster analyses. I
> have a set of data (individuals & their measurements) which I would like
> to group together. A cluster analysis will do that, but in the case of
> large numbers of individuals &/or tight groupings, I cannot read the ID
> numbers that are output to the screen (in the cluster graph). This is very
> important as I would like to be able to check a map (in a different
> program) to see if the resulting cluster analysis groups are
> geographically grouped as well.
>
> So, in short, how can I get SPSS to identify the individual IDs in each
> cluster, or at least add a new field to each individual record with a
> cluster number? Hopefuly that makes sense... Thanks,
>
> Chris
>
>
>
Reply | Threaded
Open this post in threaded view
|

Re: How to group Cluster Analysis outputs

Melissa Ives
In reply to this post by CJG
In Cluster command, use the save cluster name to keep the cluster
membership value.

[/SAVE=CLUSTER({level  })]  [/ID=varname]
                {min,max}

Melissa
The bubbling brook would lose its song if you removed the rocks.


-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
CJG
Sent: Saturday, March 10, 2007 12:10 AM
To: [hidden email]
Subject: [SPSSX-L] How to group Cluster Analysis outputs

Hello,

I am using version 11 & have a quick question about cluster analyses. I
have a set of data (individuals & their measurements) which I would like
to group together. A cluster analysis will do that, but in the case of
large numbers of individuals &/or tight groupings, I cannot read the ID
numbers that are output to the screen (in the cluster graph). This is
very important as I would like to be able to check a map (in a different
program) to see if the resulting cluster analysis groups are
geographically grouped as well.

So, in short, how can I get SPSS to identify the individual IDs in each
cluster, or at least add a new field to each individual record with a
cluster number? Hopefuly that makes sense... Thanks,

Chris


PRIVILEGED AND CONFIDENTIAL INFORMATION
This transmittal and any attachments may contain PRIVILEGED AND
CONFIDENTIAL information and is intended only for the use of the
addressee. If you are not the designated recipient, or an employee
or agent authorized to deliver such transmittals to the designated
recipient, you are hereby notified that any dissemination,
copying or publication of this transmittal is strictly prohibited. If
you have received this transmittal in error, please notify us
immediately by replying to the sender and delete this copy from your
system. You may also call us at (309) 827-6026 for assistance.