WHY 'DESCENDING' IS USED IN PROC LOGISTIC

[d6k5d3]

I was running my probit model using the following code:

proc probit data= TRY_Jump_News;
model FJ_Dummy= Abs_Surp_US1-Abs_Surp_US28 Abs_Surp_EU1-Abs_Surp_EU18 Abs_Surp_GER1-Abs_Surp_GER13 
Abs_Surp_TUR1-Abs_Surp_TUR10;
quit;

However, SAS log showed me the following warnings:

WARNING: The negative of the Hessian is not positive definite. The convergence is questionable.
WARNING: The procedure is continuing in spite of the above warning. Results shown are based on
         the last maximum likelihood iteration. Validity of the model fit is questionable

Then, I tried the following code, and I got results without any warnings:

proc logistic data= TRY_Jump_News descending;
model FJ_Dummy= Abs_Surp_US1-Abs_Surp_US28 Abs_Surp_EU1-Abs_Surp_EU18 Abs_Surp_GER1-Abs_Surp_GER13
Abs_Surp_TUR1-Abs_Surp_TUR10/ link= probit;
quit;

And I get the following notes:

NOTE: PROC LOGISTIC is modeling the probability that FJ_Dummy='1'.
NOTE: Convergence criterion (GCONV=1E-8) satisfied.
NOTE: There were 518484 observations read from the data set WORK.TRY_JUMP_NEWS.
NOTE: PROCEDURE LOGISTIC used (Total process time):
      real time           19.23 seconds
      cpu time            19.17 second

Please note the 'descending' in PROC LOGISTIC statement. I think this is what is making the difference.

Could you please explain what is the purpose of the 'descending' here? 

Much thanks.

[Reeza]

DESC

reverses the sort order for the levels of the response variable. If both the DESCENDING and ORDER= options are specified, PROC LOGISTIC orders the levels according to the ORDER= option and then reverses that order. This option has the same effect as the response variable option DESCENDING in the MODEL statement. See the section Response Level Ordering for more detail.

Run it without descending and compare the notes. 

---

@d6k5d3 wrote:

I was running my probit model using the following code:

proc probit data= TRY_Jump_News;
model FJ_Dummy= Abs_Surp_US1-Abs_Surp_US28 Abs_Surp_EU1-Abs_Surp_EU18 Abs_Surp_GER1-Abs_Surp_GER13
                Abs_Surp_TUR1-Abs_Surp_TUR10;
quit;

However, SAS log showed me the following warnings:

WARNING: The negative of the Hessian is not positive definite. The convergence is questionable.
WARNING: The procedure is continuing in spite of the above warning. Results shown are based on
         the last maximum likelihood iteration. Validity of the model fit is questionable

Then, I tried the following code, and I got results without any warnings:

proc logistic data= TRY_Jump_News descending;
model FJ_Dummy= Abs_Surp_US1-Abs_Surp_US28 Abs_Surp_EU1-Abs_Surp_EU18 Abs_Surp_GER1-Abs_Surp_GER13
                Abs_Surp_TUR1-Abs_Surp_TUR10/ link= probit;
quit;

And I get the following notes:

NOTE: PROC LOGISTIC is modeling the probability that FJ_Dummy='1'.
NOTE: Convergence criterion (GCONV=1E-8) satisfied.
NOTE: There were 518484 observations read from the data set WORK.TRY_JUMP_NEWS.
NOTE: PROCEDURE LOGISTIC used (Total process time):
      real time           19.23 seconds
      cpu time            19.17 second

Please note the 'descending' in PROC LOGISTIC statement. I think this is what is making the difference.

Could you please explain what is the purpose of the 'descending' here? 

Much thanks.

---