Johansen cointegration test

[rosemaryam]

hi
I have questions.
for J.c can i use raw data?
for estimating lag for J.c we use var s lag estimating for it.my question is about that for J.c test can i use var lag or i shoud use (var lags-1)?in some refrences i see for example if var lags were 3for J.c they use 2.

[Chuvaekek]

the data should integrated of the same order. non stationary at level and stationary in 1st difference. you could run the modified data instead to avoid confusion on lag selection

[rosemaryam]

thanks.
can u explain me.how can i modifiy data in eviews?