SPSSX Discussion

Time Series (ARIMA) question

classic Classic list List threaded Threaded
♦
♦
 5 messages Options
kienzan
Reply | Threaded
Open this post in threaded view
♦
♦
|

Time Series (ARIMA) question

kienzan
Dear List,

I've got DataSet with two columns:

+-----------+--------------------+
|  MINUTES  |  Number of errors  |
+-----------+--------------------+
|     1     |        4           |
+-----------+--------------------+
|     2     |        3           |
+-----------+--------------------+
|     3     |        2           |
+-----------+--------------------+
|   ...     |                    |


I've made both ACF (Auto Correlation Function) and PACF (Partial Auto
Correlation Function) and it seems (to me) ARIMA(0,0,1)=MA(1) model,
theta>0.

I would like to test my "hypothesis" and make some forecasts (prediction).

The best solution would be for example to show the graph of original
data and the predicted data (for an existant time period).

How can i do this?

P.S.: I'm new to SPSS and statistics so please correct my theories if
i'm wrong.

Thanks,
Alex T.

=====================
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
ViAnn Beadle
Reply | Threaded
Open this post in threaded view
♦
♦
|

Re: Time Series (ARIMA) question

ViAnn Beadle
Use the Analyze>Time Series>ARIMA dialog to specify your model and run it.
Use the Analyze>Time Series>Create Models Modeler which will run through a
variety of models and pick the best one. Help>Case Studies will walk you
through either method of performing ARIMA.

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
kienzan
Sent: Saturday, December 01, 2007 1:37 PM
To: [hidden email]
Subject: Time Series (ARIMA) question

Dear List,

I've got DataSet with two columns:

+-----------+--------------------+
|  MINUTES  |  Number of errors  |
+-----------+--------------------+
|     1     |        4           |
+-----------+--------------------+
|     2     |        3           |
+-----------+--------------------+
|     3     |        2           |
+-----------+--------------------+
|   ...     |                    |


I've made both ACF (Auto Correlation Function) and PACF (Partial Auto
Correlation Function) and it seems (to me) ARIMA(0,0,1)=MA(1) model,
theta>0.

I would like to test my "hypothesis" and make some forecasts (prediction).

The best solution would be for example to show the graph of original
data and the predicted data (for an existant time period).

How can i do this?

P.S.: I'm new to SPSS and statistics so please correct my theories if
i'm wrong.

Thanks,
Alex T.

=====================
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
kienzan
Reply | Threaded
Open this post in threaded view
♦
♦
|

Re: Time Series (ARIMA) question

kienzan
Thank you for your reply.

This Case Study "sounds" good, I've checked it.


How can i determine which is the best model for my data?


(Unfortunately I couldn't find "ARIMA" in Analyze/Time series menu.

Available menus are:

Analyze/
   Time Series/
     Create Models
     Apply Models
     --
     Seasonal Decomposition
     Spectral Analysis
     --
     Sequence Charts
     Autocorrelations
     Cross-Correlations

Can any of them replace "ARIMA" menu? What is it's aim?)



ViAnn Beadle wrote:
> Use the Analyze>Time Series>ARIMA dialog to specify your model and run it.
> Use the Analyze>Time Series>Create Models Modeler which will run through a
> variety of models and pick the best one. Help>Case Studies will walk you
> through either method of performing ARIMA.
>

=====================
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
Reutter, Alex
Reply | Threaded
Open this post in threaded view
♦
♦
|

Re: Time Series (ARIMA) question

Reutter, Alex
Analyze > Time Series > ARIMA was deprecated in favor of > Create Models.  This brings up the Time Series Modeler, which has the following modeling methods (copying and pasting from the help):
"
Expert Modeler. The Expert Modeler automatically finds the best-fitting model for each dependent series. If independent (predictor) variables are specified, the Expert Modeler selects, for inclusion in ARIMA models, those that have a statistically significant relationship with the dependent series. Model variables are transformed where appropriate using differencing and/or a square root or natural log transformation. By default, the Expert Modeler considers both exponential smoothing and ARIMA models. You can, however, limit the Expert Modeler to only search for ARIMA models or to only search for exponential smoothing models. You can also specify automatic detection of outliers.

Exponential Smoothing. Use this option to specify a custom exponential smoothing model. You can choose from a variety of exponential smoothing models that differ in their treatment of trend and seasonality.

ARIMA. Use this option to specify a custom ARIMA model. This involves explicitly specifying autoregressive and moving average orders, as well as the degree of differencing. You can include independent (predictor) variables and define transfer functions for any or all of them. You can also specify automatic detection of outliers or specify an explicit set of outliers.
"

The Expert Modeler is a good place to start; you can then tweak the model it returns by specifying a custom ARIMA model.

Cheers,
Alex


-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of kienzan
Sent: Saturday, December 01, 2007 5:30 PM
To: [hidden email]
Subject: Re: Time Series (ARIMA) question

Thank you for your reply.

This Case Study "sounds" good, I've checked it.


How can i determine which is the best model for my data?


(Unfortunately I couldn't find "ARIMA" in Analyze/Time series menu.

Available menus are:

Analyze/
   Time Series/
     Create Models
     Apply Models
     --
     Seasonal Decomposition
     Spectral Analysis
     --
     Sequence Charts
     Autocorrelations
     Cross-Correlations

Can any of them replace "ARIMA" menu? What is it's aim?)



ViAnn Beadle wrote:
> Use the Analyze>Time Series>ARIMA dialog to specify your model and run it.
> Use the Analyze>Time Series>Create Models Modeler which will run through a
> variety of models and pick the best one. Help>Case Studies will walk you
> through either method of performing ARIMA.
>

=====================
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
David Greenberg
Reply | Threaded
Open this post in threaded view
♦
♦
|

Re: Time Series (ARIMA) question

David Greenberg
In reply to this post by kienzan
When the number of counts at each time is low, as is the case here, the use of ARIMA procedures could be questioned. It assumes that the variable being modeled is continuous, and it assumes that shocks are normally distributed. If you had high counts this wouldn't be so worrisome; it would be a reasonable approximation. But here that is not the case. You may want to try Poisson regression as an alternative. David Greenberg, Sociology Department, New York University

----- Original Message -----
From: kienzan <[hidden email]>
Date: Saturday, December 1, 2007 3:45 pm
Subject: Time Series (ARIMA) question
To: [hidden email]


> Dear List,
>
> I've got DataSet with two columns:
>
> +-----------+--------------------+
> |  MINUTES  |  Number of errors  |
> +-----------+--------------------+
> |     1     |        4           |
> +-----------+--------------------+
> |     2     |        3           |
> +-----------+--------------------+
> |     3     |        2           |
> +-----------+--------------------+
> |   ...     |                    |
>
>
> I've made both ACF (Auto Correlation Function) and PACF (Partial Auto
> Correlation Function) and it seems (to me) ARIMA(0,0,1)=MA(1) model,
> theta>0.
>
> I would like to test my "hypothesis" and make some forecasts (prediction).
>
> The best solution would be for example to show the graph of original
> data and the predicted data (for an existant time period).
>
> How can i do this?
>
> P.S.: I'm new to SPSS and statistics so please correct my theories if
> i'm wrong.
>
> Thanks,
> Alex T.
>
> =====================
> 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
Free forum by Nabble Edit this page