SPSS Community News - Two New Extension Commands - STATS FIRTHLOG and STATS INEQUALITY

[Jon K Peck]

In honor of Shakespeare's birthday :-) two new extension commands have been uploaded to the SPSS Community website.

STATS FIRTHLOG estimates the Firth variation of binary logistic regression, which can address the separation issues that sometimes arise in the standard logistic model, especially when the sample size is small.

From the algorithm description...
In general, maximum likelihood estimates are often prone to small sample bias. To reduce this bias, Firth suggested to maximize the penalized log likelihood log L(beta)* = log L(beta) + 1/2 log|I(beta)|, where I(beta) is the Fisher information matrix, i. e. minus the second derivative of the log likelihood. Applying this idea to logistic regression, the score function U(beta) is replaced by the modified score function U(beta)* = U(beta) + a, where a has rth entry a_r = 0.5tr{I(beta)^{-1} [d I(beta)/d beta_r]}, r = 1,...,k.

STATS INEQUALITY calculates up to seven measures of distribution inequality for a variable, including Gini, Atkinson, Theil, and others.  It supports split files, which makes group comparisons easy.

These commands require the R Essentials and at least version 18 of Statistics.  They can be downloaded from within Statistics using the Utilities menu in version 22. For older versions you can download them from the SPSS Community website (www.ibm.com/developerworks/spssdevcentral) in the Extension Commands collection.



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

[jpatino13]

Hello Mr. Peck,

I'm attempting to add the two extension commands, Firthlogit and Stats Inequality, however I'm having a difficult time finding the extension commands. I am currently using SPSS version 21. Could you please provide me with more details to where I can retrieve the new commands.

Thank you,

Jairo

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

[Jon K Peck]

All of the extension commands are posted in the Extension Commands collection of the SPSS Community site (www.ibm.com/developerworks/spssdevcentral).  Click on Downloads for SPSS Statistics, and then look for the link to Extension Commands collection on that page.

On the collection page, you can use the search field on the upper right of the page.  Don't use the Search developerWorks but use the Search field below that.  Or you can sort the list by Name or Updated.  Setting the items per page control at the bottom of the page to the maximum facilitates going through the list as there are a lot of items posted.

HTH


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




From:        Jpatiño <[hidden email]>
To:        [hidden email], Jon K Peck/Chicago/IBM@IBMUS
Cc:        Jpatiño <[hidden email]>
Date:        08/04/2014 02:47 AM
Subject:        Re: SPSS Community News - Two New Extension Commands - STATS FIRTHLOG and STATS INEQUALITY

---

Hello Mr. Peck,

I'm attempting to add the two extension commands, Firthlogit and Stats Inequality, however I'm having a difficult time finding the extension commands. I am currently using SPSS version 21. Could you please provide me with more details to where I can retrieve the new commands.

Thank you,

Jairo


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

[jpatino13]

Thank you for the response. The information was helpful.

I've downloaded the R Package for SPSS Version 21, however Firthlogit is not running properly. I get the following messages:

GET
  FILE='C:\Users\Jpatino\Documents\Nothern_States_Dataset.sav'.
DATASET NAME DataSet1 WINDOW=FRONT.
SPSSINC BREUSCH PAGAN DEPENDENT = Claim
  ENTER = Log10_Gov_Spend
  /OPTIONS MISSING=LISTWISE
  /SAVE.
Warning message:
In readLines(file) :
  incomplete final line found on 'C:/Users/Jpatino/AppData/Local/IBM/SPSS/Statistics/21/extensions/SPSSINC_BREUSCH_PAGAN.R'
Error: The R car package is required but could not be loaded.
STATS FIRTHLOG DEPENDENT=Claim INDEP=Log10_Bud_Surp
/OPTIONS FIRTH=YES PPL=YES CONF=95
MAXIT=100 MAXSTEP=5 MAXHS=5 LCONV=.00001 XCONV=.00001.
Firth Logistic Regression: Warnings
[DataSet1] C:\Users\Jpatino\Documents\Nothern_States_Dataset.sav
Warnings
1 The R logistf package is required but could not be loaded.



Error: End of procedure


Could you please provide me with guidance in order to get FirthLogit run properly. Thank you.

Jairo

[Jon K Peck]

The first messages refer to SPSSINC BREUSCH PAGAN, not FIRTHLOG.  Older versions of R did not like a final line in the code without a carriage return.  I believe that was fixed some time ago in the BREUSCH PAGAN extension command, so you might want to download and install the current version from the SPSS Community website (In Statistics 22 packages can be installed from the Utilities menu).

However, both commands refer to a missing R package: car and logistf.  These packages would normally be installed when you install the extension command, so either you did not have an Internet connection at installation time, or the installer did not have permission to update the R library.  Try starting Statistics using Run As Administrator and repeating the bundle installs, and, of course, make sure that you have an Internet connection.


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




From:        jpatino13 <[hidden email]>
To:        [hidden email]
Date:        08/04/2014 05:50 PM
Subject:        Re: [SPSSX-L] SPSS Community News - Two New Extension Commands - STATS FIRTHLOG and STATS INEQUALITY
Sent by:        "SPSSX(r) Discussion" <[hidden email]>

---

Thank you for the response. The information was helpful.

I've downloaded the R Package for SPSS Version 21, however Firthlogit is not
running properly. I get the following messages:

GET
 FILE='C:\Users\Jpatino\Documents\Nothern_States_Dataset.sav'.
DATASET NAME DataSet1 WINDOW=FRONT.
SPSSINC BREUSCH PAGAN DEPENDENT = Claim
 ENTER = Log10_Gov_Spend
 /OPTIONS MISSING=LISTWISE
 /SAVE.
Warning message:
In readLines(file) :
 incomplete final line found on
'C:/Users/Jpatino/AppData/Local/IBM/SPSS/Statistics/21/extensions/SPSSINC_BREUSCH_PAGAN.R'
Error: The R car package is required but could not be loaded.
STATS FIRTHLOG DEPENDENT=Claim INDEP=Log10_Bud_Surp
/OPTIONS FIRTH=YES PPL=YES CONF=95
MAXIT=100 MAXSTEP=5 MAXHS=5 LCONV=.00001 XCONV=.00001.
Firth Logistic Regression: Warnings
[DataSet1] C:\Users\Jpatino\Documents\Nothern_States_Dataset.sav
Warnings                
1                 The R logistf package is required but could not be loaded.



Error: End of procedure


Could you please provide me with guidance in order to get FirthLogit run
properly. Thank you.

Jairo



--
View this message in context: http://spssx-discussion.1045642.n5.nabble.com/SPSS-Community-News-Two-New-Extension-Commands-STATS-FIRTHLOG-and-STATS-INEQUALITY-tp5725639p5726883.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

[stinehoj]

Am I correct in believing that STATS FIRTHLOG does not support split files? I am attempting to run a loigistic regression on 1000 bootstrap samples contained within a single dataset. It seems to return estimates for the first sample only.