EViews.com

Posted: **Mon Jun 26, 2017 8:56 am**

Hi,

I am having a problem with extracting the results from tables (in particular RESET test). I am using stepls in a loop and as expected each regression has a different number of independent variables. The problem is that the row number in the frozen table containing the statistic (or the p-value) of the RESET test keeps changing depending on the number of regressors that are included. Does anyone know how to solve this? is there another way to extract the relevant statistics from the tests?

This is the code I am using:

Code: Select all

!size=@obs(complete)
!regressions=@obs(complete)
for !y= 1 to !regressions
!x=complete(!y)
'Ramsey RESET 
for !a=1 to !resettest
freeze(rtable) temp{!x}.reset(!a)       
vector (!y) rt{!a}  
rt{!a}(!y)=rtable(9,4)
 d rtable
next
next

Many thanks!

Francisco

Posted: **Mon Jun 26, 2017 9:02 am**

Seems like the p-value is always in column D, and is going to be in row 7,8,9, etc... depending on how many regressors there are (and the length of the names).

Since there are no number above the p-value in column D, you could just extract D7, if it isn't a number, extract D8, and keep going until you get a number.

Alternatively, the reset test is pretty easy to run manually by doing the regression.

Posted: **Mon Jun 26, 2017 4:40 pm**

Thanks, solution found!

Code: Select all

!resettest=4

!size=@obs(complete)
!regressions=@obs(complete)
for !y= 1 to !regressions
!x=complete(!y)
'Ramsey RESET 
for !a=1 to !resettest
freeze(rtable) temp{!x}.reset(!a)   
for !e=7 to 12
!xx2=@isna(@val(rtable(!e,4)))
if !xx2=0 then
exitloop
endif
next    
vector (!y) rt{!a}  
rt{!a}(!y)=rtable(!e,4)
 d rtable
next
next

Posted: **Mon Jun 26, 2017 6:21 pm**

Nicely done.

Posted: **Mon Jun 26, 2017 11:54 pm**

Thanks!!

EViews.com

Extracting results from tables

Extracting results from tables

Re: Extracting results from tables

Re: Extracting results from tables

Re: Extracting results from tables

Re: Extracting results from tables