CLASSIFICATION

[Peter Spangler]

Using the SPSS APRIORI INC command is it possible to then assign a
value (yes/no 1/0) for those cases that meet the "rule" in the
classification table created in the output.

Peter

Sent from my iPhone

=====================
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

[Jon K Peck]

You will have many rules in the output table, so this question could mean several things including:
- create an indicator variable for each rule indicating whether a case meets it
- create an indicator variable for each rule meeting some support and confidence thresholds and create indicator variables for those
- select a particular consequent and calculate on the estimation or new data which cases meet it.

In order to do something like this, you would capture the table with OMS parse the rules column and generate the requisite transformation.  This would be most easily done with a little Python code.  If you would spell out the actual task, I can help with that.


Jon Peck (no "h") aka Kim
Senior Software Engineer, IBM
[hidden email]
phone: [phone]




From:        Peter Spangler <[hidden email]>
To:        [hidden email],
Date:        08/23/2013 10:51 PM
Subject:        [SPSSX-L] CLASSIFICATION
Sent by:        "SPSSX(r) Discussion" <[hidden email]>

---

Using the SPSS APRIORI INC command is it possible to then assign a
value (yes/no 1/0) for those cases that meet the "rule" in the
classification table created in the output.

Peter

Sent from my iPhone

=====================
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

[Peter Spangler]

 create an indicator variable for each rule meeting some support and confidence thresholds and create indicator variables for those 

YES, I would like to create indicator variables where 

Rule A: {popcorn, movie}=>{beer} with support .01 and confidence .40 

- select a particular consequent and calculate on the estimation or new data which cases meet it. 

YES, estimate which new cases will meet Rule A.

Using the Python module sounds like a great implementation resource as well! 

Sent from my iPhone

On Aug 24, 2013, at 5:43 AM, Jon K Peck <[hidden email]> wrote:

You will have many rules in the output table, so this question could mean several things including:
- create an indicator variable for each rule indicating whether a case meets it
- create an indicator variable for each rule meeting some support and confidence thresholds and create indicator variables for those
- select a particular consequent and calculate on the estimation or new data which cases meet it.

In order to do something like this, you would capture the table with OMS parse the rules column and generate the requisite transformation.  This would be most easily done with a little Python code.  If you would spell out the actual task, I can help with that.


Jon Peck (no "h") aka Kim
Senior Software Engineer, IBM
[hidden email]
phone: [phone]




From:        Peter Spangler <[hidden email]>
To:        [hidden email],
Date:        08/23/2013 10:51 PM
Subject:        [SPSSX-L] CLASSIFICATION
Sent by:        "SPSSX(r) Discussion" <[hidden email]>

---

Using the SPSS APRIORI INC command is it possible to then assign a
value (yes/no 1/0) for those cases that meet the "rule" in the
classification table created in the output.

Peter

Sent from my iPhone

=====================
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