Singular matrix
Posted: Mon Dec 18, 2017 11:31 pm
Hi
Not sure this is a bug, but the error message should be different. When i run the regression:
I get the error message
I believe this is caused by the variable KWA90 beginning in 1992 and therefore Eviews does not like it when i include a dummy variable for 1987q1 (@DURING("87Q1")). But, ideally, Eviews should just ignore such a dummy?
Thomas
Not sure this is a bug, but the error message should be different. When i run the regression:
Code: Select all
EQUATION _EQ_WW90I.LS DLOG(WW90I) = C(1)+C(2)*DLOG(WW90I(-2))+C(3)*DLOG(WW90I(-3))+C(4)*LOG(WW90I(-1)/KWA90(-1))+C(5)*DLOG(KWA90)+C(6)*LOG(URKORR(-1))+C(8)*@DURING("87Q1")I get the error message
- nearsingularmatrixerror.PNG (9.77 KiB) Viewed 3200 times
I believe this is caused by the variable KWA90 beginning in 1992 and therefore Eviews does not like it when i include a dummy variable for 1987q1 (@DURING("87Q1")). But, ideally, Eviews should just ignore such a dummy?
Thomas