SPSS Forecasting Module

[Stan Gorodenski]

Some time ago I was lamenting the fact that, at least 7 years ago,
SPSS's ARIMA procedure was quite limited. It could not do multivariate
ARIMA models, and one could not control the inclusion of orders in
univariate models. In other words, if I had a 3rd order autoregressive
univariate model and the 2nd order was not significant, I could not
exclude it. SPSS required all three orders in the final model. There was
no way to remove the 2nd order. Someone on this list asked if I had
tried xxx, where xxx is the name of some kind of SPSS procedure, I
think, that I do not remember. Could someone tell me again? Maybe it is
part of the SPSS Forecasting module

I am interested in SPSS's Forecasting module, primarily with regard to
the ARIMA and SPECTRA part of it. Does anyone know if it has the same
flexibility as the PACK program? David Pack in the 1970's wrote a
program under a government grant that did univariate and multivariate
ARIMA models. In the 1980's I attended a work related workshop on ARIMA
forecasting. The instructor gave all attendees a free copy of the PACK
program. To my surprise, the state department I worked in installed it
on the mainframe. With PACK I could control completely the model I was
developing, and exclude orders that were not significant. The output had
all the significance levels for model building. Do the ARIMA and SPECTRA
procedures also have this same flexibility? Can one build models by
writing syntax and submitting batch programs (on the pc) instead of
using drop down menus? It appears in the white paper on the Forecasting
module that it has a lot of capabilities (excluding auto model building
which I do not believe in), but the white paper might be misleading.
Stan

=====================
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]

Do a search on ibm spss statistics TSMODEL. It should turn up some
online help.

<a href="http://publib.boulder.ibm.com/infocenter/spssstat/v20r0m0/index.jsp?topic=%2">http://publib.boulder.ibm.com/infocenter/spssstat/v20r0m0/index.jsp?topic=%2
Fcom.ibm.spss.statistics.help%2Fsyn_tsmodel.htm

You should also be able to find the downloadable pdf for the
forecasting option.

Tony Babinec
[hidden email]


-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Stan Gorodenski
Sent: Thursday, June 20, 2013 3:35 PM
To: [hidden email]
Subject: SPSS Forecasting Module

Some time ago I was lamenting the fact that, at least 7 years ago, SPSS's
ARIMA procedure was quite limited. It could not do multivariate ARIMA
models, and one could not control the inclusion of orders in univariate
models. In other words, if I had a 3rd order autoregressive univariate model
and the 2nd order was not significant, I could not exclude it. SPSS required
all three orders in the final model. There was no way to remove the 2nd
order. Someone on this list asked if I had tried xxx, where xxx is the name
of some kind of SPSS procedure, I think, that I do not remember. Could
someone tell me again? Maybe it is part of the SPSS Forecasting module

I am interested in SPSS's Forecasting module, primarily with regard to the
ARIMA and SPECTRA part of it. Does anyone know if it has the same
flexibility as the PACK program? David Pack in the 1970's wrote a program
under a government grant that did univariate and multivariate ARIMA models.
In the 1980's I attended a work related workshop on ARIMA forecasting. The
instructor gave all attendees a free copy of the PACK program. To my
surprise, the state department I worked in installed it on the mainframe.
With PACK I could control completely the model I was developing, and exclude
orders that were not significant. The output had all the significance levels
for model building. Do the ARIMA and SPECTRA procedures also have this same
flexibility? Can one build models by writing syntax and submitting batch
programs (on the pc) instead of using drop down menus? It appears in the
white paper on the Forecasting module that it has a lot of capabilities
(excluding auto model building which I do not believe in), but the white
paper might be misleading.
Stan

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

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