i have made some modification for the program in this page http://forums.eviews.com/viewtopic.php?f=5&t=1638, i would like to estimate the best pdl(y,!i,!j) by loop that minimise AIC, but when i use this code i get error message " Near singular matrix error. Regressors may be perfectly
collinear in "DO_ EQ.LS Y C PDL(Y,1,2)"; i have estimate all this models manually but i can not get this error!
Code: Select all
' open workfile
wfopen C:\Users\MED\Documents\quandl_seas.WF1
pageselect quarterly
'create empty equation to be used inside the loop
equation eq
'variable to store the minimum AIC. Initialise it to a large number
!aic = 99999999
'variable saying how many lags and order to go up to
!maxlags = 12
!maxorders= 5
'Variable to store the "best" number of lags and order
!bestlag = 0
!bestorder=0
'set sample to be the !maxlag'th value onwards
smpl @first+!maxlags @last
for !i=1 to !maxlags
for !j=1 to !maxorders
eq.ls y c pdl(y,!i,!j) 'run pdl of Y on a constant with different lag & order
if eq.@aic < !aic then
!bestlag = !i 'if this lag specification has the best AIC, then store this lag as !bestlag.
!bestorders=!j 'if this order specification has the best AIC, then store this lag as !bestorder.
!aic = eq.@aic
endif
next
next
show eq.ls y c pdl(y,!bestlags,!bestorders)
'reset sample
smpl @all