Syntax to obtain variable information about cluster membership

[bgreen]

Dear all,

In 'IBM SPSS Statistics 19 Statistical Procedures Companion' Norusis
noted: "To see how clusters differ on the variables used to create them,
save the cluster membership number using the Save command and then use the
Means procedure, specifying the variables used to form the clusters as the
dependent variables and the cluster number as the grouping variable"

I have two questions:

1) Is it possible to obtain information about the variables when the
variables are binary?

2) if so, I was hoping someone could advise me regarding the syntax to
obtain information about cluster composition, assuming variables var1 ....
var11). I assume the use of the means command would not apply for binary
data.

Any assistance is appreciated,

Bob

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Melissa Ives]

Bob,

If the 'binary' variables are represented as 0 or 1, the mean is the same as the percent endorsing (answer=1) the variables, so Norusis' recommendation works well.
Means do not work for nominal (categorical data).

Melissa

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Bob Green
Sent: Wednesday, June 20, 2012 12:34 AM
To: [hidden email]
Subject: [SPSSX-L] Syntax to obtain variable information about cluster membership

Dear all,

In 'IBM SPSS Statistics 19 Statistical Procedures Companion' Norusis
noted: "To see how clusters differ on the variables used to create them, save the cluster membership number using the Save command and then use the Means procedure, specifying the variables used to form the clusters as the dependent variables and the cluster number as the grouping variable"

I have two questions:

1) Is it possible to obtain information about the variables when the variables are binary?

2) if so, I was hoping someone could advise me regarding the syntax to obtain information about cluster composition, assuming variables var1 ....
var11). I assume the use of the means command would not apply for binary data.

Any assistance is appreciated,

Bob

=====================
To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

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 [phone] for assistance.

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[bgreen]

Hello Melissa,

Many thanks for confirming this. Is it possible to perform Norusis'
recommendation through syntax? I don't have access to SPSS at home
but am not sure how the cluster membership number would be identified
as such in syntax?

Regards

Bob



At 09:19 PM 20/06/2012, you wrote:

>Bob,
>
>If the 'binary' variables are represented as 0 or 1, the mean is the
>same as the percent endorsing (answer=1) the variables, so Norusis'
>recommendation works well.
>Means do not work for nominal (categorical data).
>
>Melissa
>
>-----Original Message-----
>From: SPSSX(r) Discussion [mailto:[hidden email]] On
>Behalf Of Bob Green
>Sent: Wednesday, June 20, 2012 12:34 AM
>To: [hidden email]
>Subject: [SPSSX-L] Syntax to obtain variable information about
>cluster membership
>
>Dear all,
>
>In 'IBM SPSS Statistics 19 Statistical Procedures Companion' Norusis
>noted: "To see how clusters differ on the variables used to create
>them, save the cluster membership number using the Save command and
>then use the Means procedure, specifying the variables used to form
>the clusters as the dependent variables and the cluster number as
>the grouping variable"
>
>I have two questions:
>
>1) Is it possible to obtain information about the variables when the
>variables are binary?
>
>2) if so, I was hoping someone could advise me regarding the syntax
>to obtain information about cluster composition, assuming variables var1 ....
>var11). I assume the use of the means command would not apply for binary data.
>
>Any assistance is appreciated,
>
>Bob
>
>=====================
>To manage your subscription to SPSSX-L, send a message to
>[hidden email] (not to SPSSX-L), with no body text except
>the command. To leave the list, send the command SIGNOFF SPSSX-L For
>a list of commands to manage subscriptions, send the command INFO REFCARD
>
>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 [phone] for assistance.

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Melissa Ives]

From the Syntax Help in SPSS:

SAVE Subcommand (CLUSTER command)
SAVE allows you to save cluster membership at specified solution levels as new variables in the active dataset.

* The specification on SAVE is the CLUSTER keyword, followed by either a single number indicating the level (number of clusters) of the cluster solution or a range separated by a comma indicating the minimum and maximum numbers of clusters when membership of more than one solution is to be saved. The number or range must be enclosed in parentheses and applies to all methods specified on METHOD.

* You can specify a rootname in parentheses after each method specification on the METHOD subcommand. CLUSTER forms new variable names by appending the number of the cluster solution to the rootname.

* If no rootname is specified, CLUSTER forms variable names using the formula CLUn_m, where m increments to create a unique rootname for the set of variables saved for one method and n is the number of the cluster solution.

* The names and descriptive labels of the new variables are displayed in the procedure information notes.

* You cannot use the SAVE subcommand if you are replacing the active dataset with matrix materials (See the topic Matrix Output (CLUSTER command) for more information. )

Example
CLUSTER A B C
 /METHOD=BAVERAGE SINGLE (SINMEM) WARD
 /SAVE=CLUSTERS(3,5).
* This command creates nine new variables: CLU5_1, CLU4_1, and CLU3_1 for BAVERAGE, SINMEM5, SINMEM4, and SINMEM3 for SINGLE, and CLU5_2, CLU4_2, and CLU3_2 for WARD. The variables contain the cluster membership for each case at the five-, four-, and three-cluster solutions using the three clustering methods. Ward's method is the third specification on METHOD but uses the second set of default names, since it is the second method specified without a rootname.

* The order of the new variables in the active dataset is the same as listed above, since the solutions are obtained in the order from 5 to 3.

* New variables are listed in the procedure information notes.


-----Original Message-----
From: Bob Green [mailto:[hidden email]]
Sent: Wednesday, June 20, 2012 6:56 AM
To: Melissa Ives; [hidden email]
Subject: RE: [SPSSX-L] Syntax to obtain variable information about cluster membership

Hello Melissa,

Many thanks for confirming this. Is it possible to perform Norusis'
recommendation through syntax? I don't have access to SPSS at home but am not sure how the cluster membership number would be identified as such in syntax?

Regards

Bob



At 09:19 PM 20/06/2012, you wrote:

>Bob,
>
>If the 'binary' variables are represented as 0 or 1, the mean is the
>same as the percent endorsing (answer=1) the variables, so Norusis'
>recommendation works well.
>Means do not work for nominal (categorical data).
>
>Melissa
>
>-----Original Message-----
>From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf
>Of Bob Green
>Sent: Wednesday, June 20, 2012 12:34 AM
>To: [hidden email]
>Subject: [SPSSX-L] Syntax to obtain variable information about cluster
>membership
>
>Dear all,
>
>In 'IBM SPSS Statistics 19 Statistical Procedures Companion' Norusis
>noted: "To see how clusters differ on the variables used to create
>them, save the cluster membership number using the Save command and
>then use the Means procedure, specifying the variables used to form the
>clusters as the dependent variables and the cluster number as the
>grouping variable"
>
>I have two questions:
>
>1) Is it possible to obtain information about the variables when the
>variables are binary?
>
>2) if so, I was hoping someone could advise me regarding the syntax to
>obtain information about cluster composition, assuming variables var1 ....
>var11). I assume the use of the means command would not apply for binary data.
>
>Any assistance is appreciated,
>
>Bob
>
>=====================
>To manage your subscription to SPSSX-L, send a message to
>[hidden email] (not to SPSSX-L), with no body text except
>the command. To leave the list, send the command SIGNOFF SPSSX-L For a
>list of commands to manage subscriptions, send the command INFO REFCARD
>
>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 [phone] for
>assistance.

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 [phone] for assistance.

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD