Re: VERY IMPORTANT.....SPSS multinomial logistic regression

On Jun 19, 4:26 pm, "A.A.A" <ayaf...@xxxxxxxxx> wrote:
Hi,
I am making an SPSS syntax that should take data written on an excel
file, then apply a multinomail logit model to this
data. This syntax should be repeated for 50
datasets(datas1,datas2,...........datas50).I want to get 2 outputs

1-the estimated parameters
2-the percentage of correct classification

But i want this output to be saved in excel files(not in the form of
tables as the SPSS output)
so that
         output file "1" carries the parameters and % of correct
classification resulting from applying the multinomial logit to
datas1
          output file "2" carries the parameters and % of correct
classification resulting from applying the multinomial logit to
datas2
          output file "3" carries the parameters and % of correct
classification resulting from applying the multinomial logit to
datas3
 and so on till datas50

First ,how could i run the multinomial logit through the syntax?
second,how could i run this 50 times?
third, how could i get the output as desired?

Would it be like this?if so, how to complete with this syntax?

*loading the excel file containing the data.
*****************************************************.
Define !Xel().
!do !i=1 !to 50.
GET DATA
  /TYPE=XLS
  /FILE=!Quote(!concat('E:\simulationexcel\', 'datas', !i, '.xls'))
  /SHEET=name 'Sheet1'
  /CELLRANGE=full
  /READNAMES=on
  /ASSUMEDSTRWIDTH=32767.

 *saving the file after droping unwanted columns, then loading the
new
file.
***************************************************************************­­­
********.
SAVE OUTFILE= !quote(!concat('E:\work\', 'datas', !i, '.sav'))
  /DROP=V1 V4 V5 V6 V7 V8 V9 V10 V11 V12 V13 /COMPRESSED.

GET
  FILE= !quote(!concat('E:\work\', 'datas', !i, '.sav')).
DATASET NAME DataSet3 WINDOW=FRONT.

*computing MLE.
******************

NOMREG
  V14  (BASE=LAST ORDER=ASCENDING) WITH @2 @1
  /CRITERIA CIN(95) DELTA(0) MXITER(100) MXSTEP(5) CHKSEP(20)
LCONVERGE(0) PCONVERGE(0.000001) SINGULAR(0.00000001)
  /MODEL
  /STEPWISE = PIN(.05) POUT(0.1) MINEFFECT(0) RULE(SINGLE)
ENTRYMETHOD(LR) REMOVALMETHOD(LR)
  /INTERCEPT =INCLUDE
  /PRINT = CLASSTABLE PARAMETER SUMMARY LRT CPS STEP MFI
  /SAVE PREDCAT .

Waiting for your reply
Thanks inadvance

Are you paying me? Or do you just sound like you are?

illywhacker;
.

Loading