EViews.com

Hi
I want to write a program and get IRFs in a VAR model by 3 lags.i have 6 variables and 97 observation

i write:

var var_01.ls 1 3 I_obs Y_obs G_obs O_obs M_obs inf_obs
var_01.impulse(20,m,se=a,matbys=imp_var) I_obs Y_obs G_obs O_obs M_obs inf_obs @ M_obs
But this error occurs:
Insufficient number of observation to estimate 19 coefficients per equation.

I increased the number of observations but Error still exists. what should i do?

Thanks for your help.

You probably don't have 97 valid observations. Check the series for NAs. If that doesn't work, post your workfile.