SPSS Statistics 17 Extension Commands VII - Apriori (Rules) and Fancy Correlations

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
|

SPSS Statistics 17 Extension Commands VII - Apriori (Rules) and Fancy Correlations

Peck, Jon
We're getting close to the end of this series - at least for now.  New extension commands will doubtless keep coming along from SPSS and/or from users.  The things that can be done are endless.

 

For today, I'll mention two more R-based extension commands:

 

SPSSINC APRIORI, and

SPSSINC HETCOR.

 

SPSSINC APRIORI finds association rules in transactions such as if you buy products a, b, and q, you are likely to buy w.  It is built on the R arules package.  It's classified as a machine learning tool in the R task views.  It might complement the new RFM option in SPSS Statistics 17 (not an extension command), but I'm on unfamiliar terrain here.

 

SPSSINC HETCOR calculates correlations (what, yet another correlation command?).  The type of the correlation calculated depends on the measurement levels of the pair of variables correlated.  Besides ordinary Pearson correlation between scale variables, it calculates polychoric and polyserial correlations when nominal or ordinal variables are involved.  The output includes a table that shows what type of correlation was computed for each pair.  It is built on the R Hetcor package.

 

Both of these commands have dialog boxes (with help) as well as traditional style SPSS Statistics syntax.

 

Both of these packages will be posted soon to Developer Central.

 

 

Jon K. Peck

SPSS Inc.

[hidden email] <mailto:[hidden email]>

(ip) phone 312-651-3435

 

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