Re: [Gretl-users] issue

Dear Sven and Allin, Thanks for the reply. I am not sure whether I understood your comments correctly. Let me post some more stuff of the code: ---------------------------------------------- function matrix mplier(scalar a, matrix b, matrix varphi, matrix vpi, scalar l, scalar h, scalar k) vphi = zeros(1,l) vphi[1] = 1 + a + varphi[1] loop for (i=2; i=l-1; i+=1) vphi[i] = varphi[i] - varphi[i-1] endloop vphi[l] = -varphi[l-1] mtheta = zeros(l+1,k) mtheta[1,] = vpi[1,] mtheta[2,] = vpi[2,] - vpi[1,] + b loop for (i=3; i=l; i+=1) mtheta[i,] = vpi[i,] - vpi[i-1,] endloop mtheta[l+1,] = -vpi[l-1,] mpsi = zeros(h,k) mpsi[1,] = mtheta[1,] loop for (i=1; i=l; i+=1) mpsi[1+i,] = vphi[1:i]*mpsi[i:1,] + mtheta[1+i,] endloop mpsi[1+1,] = vphi[1:1]*mpsi[1:1,] + mtheta[1+1,] #run the loop manually mpsi[2+1,] = vphi[1:2]*mpsi[2:1,] + mtheta[2+1,] #run the loop manually mpsi[3+1,] = vphi[1:3]*mpsi[3:1,] + mtheta[3+1,] #run the loop manually loop for (i=l+1; i=h-1; i+=1) mpsi[i+1,] = vphi*mpsi[i:i-l+1,] endloop end function ----------------------------------------------- The first two blocks run fine and I obtain the same results as in gauss. The third block makes trouble. The corresponding Gauss code - which works fine - for the third block is: GAUSS------------------------------- mpsi = zeros(h,k) ; mpsi[1,.] = mtheta[1,.] ; for i (1,l,1) ; mpsi[i+1,.] = vphi[1:i]*mpsi[i:1,.] + mtheta[i+1,.] ; endfor ; for i (l+1,h-1,1) ; mpsi[i+1,.] = vphi*mpsi[i:i-l+1,.] ; endfor ; ---------------------------------------- There I dont obtain an error running the loop. I am really not sure how to translate the GAUSSian code lines into proper gretl-language. The mreverse command did not help really. Best, Artur