SPSSX Discussion

Re: ver 15 error message -never mind

Classic

List

Threaded

4 messages Options

Raffe, Sydelle, SSA

Re: ver 15 error message -never mind

I should have tried closing and opening SPSS before I posted my question. There my file was. Thanks anyway.

Jarrod Teo-2

IBM SPSS Modeler v14.1

Hi,

I am working on a project which requires Apriori.

When I run the model, the error says that the threshold is too high and no models were generated. Any idea why is this happening? I am running on a data file of 500K rows with 81 columns using Apriori.

My data file for Apriori is as such:

id             prod1 prod2 prod3
1                1        0        1
2                1        0        1
3                0        1        1
.
.
.
.
.
500,000

By the way, I run with 750 rows and 10 columns, and this error appears as well for Apriori.

Thanks for any help.

Warmest Regards
Dorraj Oet

MaxJasper

Re: IBM SPSS Modeler v14.1

Rearrange your data file as follows:

Name your 80 products from 1,2,.,80

Then data files looks like:

Id product sets
1 1,4,76,....
2 33, 21, 71, .....
3 65,11,7, 80,....

Max.

From: DorraJ Oet [via SPSSX Discussion]
[mailto:[hidden email]]
Sent: Monday, October 10, 2011 20:48
To: MaxJasper
Subject: IBM SPSS Modeler v14.1

Hi,

I am working on a project which requires Apriori.

When I run the model, the error says that the threshold is too high and no
models were generated. Any idea why is this happening? I am running on a
data file of 500K rows with 81 columns using Apriori.

My data file for Apriori is as such:

id prod1 prod2 prod3
1 1 0 1
2 1 0 1
3 0 1 1
.
.
.
.
.
500,000

By the way, I run with 750 rows and 10 columns, and this error appears as
well for Apriori.

Thanks for any help.

Warmest Regards
Dorraj Oet

________________________________________
If you reply to this email, your message will be added to the discussion
below:
http://spssx-discussion.1045642.n5.nabble.com/Re-ver-15-error-message-never-
mind-tp4890158p4890511.html
To start a new topic under SPSSX Discussion, email
[hidden email]
To unsubscribe from SPSSX Discussion, click here.

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

Anthony Babinec

Re: IBM SPSS Modeler v14.1

In reply to this post by Jarrod Teo-2

On the TYPE node, make sure the fields are set to a role of BOTH.

On the APRIORI node, Model tab, check the box “Only true value for flags.”

Understand what the message is saying. You have 3 setable tuning parameters:

support (default 10%), confidence (default 80), max number of antecedents (default 5).

Depending on the patterns in the data, it is possible that the above settings

return no rules. You might need to lower either the confidence or the support

settings or both. In particular, if items are not very frequent, then support of 10

could be too high.

You might explore your fields using the Distribution node to learn how frequent the

1s (yes values) are. That will give you a sense of what Support value to use.

Also – I am not sure of this – but I believe it the case that Support can go no

lower than 1. If this is so, it would be nice to see Apriori changed so that

lower supports are allowed.

Tony Babinec

From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of DorraJ Oet
Sent: Monday, October 10, 2011 9:34 PM
To: [hidden email]
Subject: IBM SPSS Modeler v14.1