SET WORKSPACE Issue

[The Burl]

Greetings,

Version: 22
Objective: Trying to run a MVA Little's test on a very large dataset

Issue: Instructed to SET WORKSPACE = 689343. I do as instructed and receive the identical warning after trying to run it again.

Thanks, as always.
Jon

[David Marso]

Maybe you need to be a bit more specific?
What do you define as 'very large'?
Some people think 1000 or 10000 cases is vary large.
Some people consider that 10,000,000 cases is routine.
Post your syntax?
How many variables are you using? Are they Scale or categorical?
Have you tried setting WORKSPACE to an even larger number than 689343?
What version of Windoze?  32 or 64 bit? Same with SPSS.
These are the things someone probably needs to know in order to make any sort of recommendations.

The Burl wrote

Greetings,

Version: 22
Objective: Trying to run a MVA Little's test on a very large dataset

Issue: Instructed to SET WORKSPACE = 689343. I do as instructed and receive the identical warning after trying to run it again.

Thanks, as always.
Jon

[The Burl]

MVA VARIABLES=Communication Feedback TeamworkWithin ManagerExpec OrgLearning ManageSupport
    OverallPerceptions FrequencyReports TeamworkAcross Staffing Handoffs Nonpunitive
  /EM(TOLERANCE=0.001 CONVERGENCE=0.0001 ITERATIONS=10).

About 300,000 cases.

I've run SET WORKSPACE = 1000000. too to no avail.
Thank you David in advance!

[The Burl]

I think it's behaving now, got it to run using one variable at a time.

However, the Little's MCAR test results don't make sense:
 "Little's MCAR test: Chi-Square = .000, DF = 0, Sig. = ."

Syntax is:
MVA VARIABLES=D1
  /EM(TOLERANCE=0.001 CONVERGENCE=0.0001 ITERATIONS=25).


Any ideas? Anyone run into this before?

[David Marso]

I know next to nothing about MVA and don't even have the option installed.
However, IMNSHO it makes zero intuitive sense to run it on a single variable.
Think about it.
--

The Burl wrote

I think it's behaving now, got it to run using one variable at a time.

However, the Little's MCAR test results don't make sense:
 "Little's MCAR test: Chi-Square = .000, DF = 0, Sig. = ."

Syntax is:
MVA VARIABLES=D1
  /EM(TOLERANCE=0.001 CONVERGENCE=0.0001 ITERATIONS=25).


Any ideas? Anyone run into this before?

[Jon K Peck]

With only one variable, the imputed value from MVA would typically just be the mean FWIW.


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




From:        David Marso <[hidden email]>
To:        [hidden email]
Date:        08/03/2015 10:46 AM
Subject:        Re: [SPSSX-L] SET WORKSPACE Issue
Sent by:        "SPSSX(r) Discussion" <[hidden email]>

---

I know next to nothing about MVA and don't even have the option installed.
However, IMNSHO it makes zero intuitive sense to run it on a single
variable.
Think about it.
--

The Burl wrote

> I think it's behaving now, got it to run using one variable at a time.
>
> However, the Little's MCAR test results don't make sense:
>  "Little's MCAR test: Chi-Square = .000, DF = 0, Sig. = ."
>
> Syntax is:
> MVA VARIABLES=D1
>   /EM(TOLERANCE=0.001 CONVERGENCE=0.0001 ITERATIONS=25).
>
>
> Any ideas? Anyone run into this before?

-----
Please reply to the list and not to my personal email.
Those desiring my consulting or training services please feel free to email me.
---
"Nolite dare sanctum canibus neque mittatis margaritas vestras ante porcos ne forte conculcent eas pedibus suis."
Cum es damnatorum possederunt porcos iens ut salire off sanguinum cliff in abyssum?"
--
View this message in context: http://spssx-discussion.1045642.n5.nabble.com/SET-WORKSPACE-Issue-tp5730325p5730356.html
Sent from the SPSSX Discussion mailing list archive at Nabble.com.

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