gauss loop

/**     Compute variance decompositions for horizons 1 to 10    **/
vd1  = ir1.^2;
vd1  = 100*vd1./sumc(vd1');
vd2  = ir1.^2 + ir2.^2;
vd2  = 100*vd2./sumc(vd2');
vd3  = ir1.^2 + ir2.^2 + ir3.^2;
vd3  = 100*vd3./sumc(vd3');
vd4  = ir1.^2 + ir2.^2 + ir3.^2 + ir4.^2;
vd4  = 100*vd4./sumc(vd4');
vd5  = ir1.^2 + ir2.^2 + ir3.^2 + ir4.^2 + ir5.^2;
vd5  = 100*vd5./sumc(vd5');
vd6  = ir1.^2 + ir2.^2 + ir3.^2 + ir4.^2 + ir5.^2 + ir6.^2;
vd6  = 100*vd6./sumc(vd6');
vd7  = ir1.^2 + ir2.^2 + ir3.^2 + ir4.^2 + ir5.^2 + ir6.^2 + ir7.^2;
vd7  = 100*vd7./sumc(vd7');
vd8  = ir1.^2 + ir2.^2 + ir3.^2 + ir4.^2 + ir5.^2 + ir6.^2 + ir7.^2 + ir8.^2;
vd8  = 100*vd8./sumc(vd8');
vd9  = ir1.^2 + ir2.^2 + ir3.^2 + ir4.^2 + ir5.^2 + ir6.^2 + ir7.^2 + ir8.^2 + ir9.^2;
vd9  = 100*vd9./sumc(vd9');
vd10 = ir1.^2 + ir2.^2 + ir3.^2 + ir4.^2 + ir5.^2 + ir6.^2 + ir7.^2 + ir8.^2 + ir9.^2 + ir10.^2;
vd10 = 100*vd10./sumc(vd10');

------------------------------------------------------------------------------------------------------------------

 

then I want to sum

sumc(diag(vd1*vd1') to   sumc(diag(vd10*vd10')

so I use do until.

 

tr=0;
i=1;
do until i>10;
tr = tr + sumc(diag(vd[i]*vd[i]'));
i=i+1;
endo;

 

but error occur...

please help me.

 

 

5 Answers



0



You are getting an error, because inside of your do loop you are trying to index into a matrix named vd. However, that matrix does not exist. There is a way to loop through global variables like you are wanting to do, but I think we can make the code much shorter and simpler if we know a few things about the data. Do all of the ir matrices have the same dimensions? What are they?

The simplest, though not most elegant, solution is to simply replace your do loop with this:

tr=0;
i=1;
do until i>10;
    tmp = varget("vd"$+ntos(i));
    tr = tr + sumc(diag(tmp*tmp'));
    i=i+1;
endo;

One step better would be to simplify by using a for loop in place of the do loop like this:

for i(1, 10, 1);
    tmp = varget("vd"$+ntos(i));
    tr = tr + sumc(diag(tmp*tmp'));
endfor;

But if you answer the question from above about the sizes of these variables, we can make a version that will be more elegant and you will not have to hand enter each of the terms. You will appreciate this if you change your model to include more.

aptech

1,773


0



I am user gauss version 10.

and I think the ntos(i) is for gauss version 14.

would you revise your coding for gauss 10 version?

It'd be very appreciate that.

lee

0


0



I resolve the problem using cvtos(ftocv(i, 0, 0));

thnaks you.

lee

0


0



and  all of the ir matrices have the same dimensions.

please let me know how I convert my code more elegan.

lee

0


0



Yes, GAUSS 10 does not have the function ntos. For older versions of GAUSS replace:

tmp = varget("vd"$+ntos(i));

with:

tmp = varget("vd"$+cvtos(ftocv(i, 1, 0)));

aptech

1,773

Your Answer

5 Answers

0

You are getting an error, because inside of your do loop you are trying to index into a matrix named vd. However, that matrix does not exist. There is a way to loop through global variables like you are wanting to do, but I think we can make the code much shorter and simpler if we know a few things about the data. Do all of the ir matrices have the same dimensions? What are they?

The simplest, though not most elegant, solution is to simply replace your do loop with this:

tr=0;
i=1;
do until i>10;
    tmp = varget("vd"$+ntos(i));
    tr = tr + sumc(diag(tmp*tmp'));
    i=i+1;
endo;

One step better would be to simplify by using a for loop in place of the do loop like this:

for i(1, 10, 1);
    tmp = varget("vd"$+ntos(i));
    tr = tr + sumc(diag(tmp*tmp'));
endfor;

But if you answer the question from above about the sizes of these variables, we can make a version that will be more elegant and you will not have to hand enter each of the terms. You will appreciate this if you change your model to include more.

0

I am user gauss version 10.

and I think the ntos(i) is for gauss version 14.

would you revise your coding for gauss 10 version?

It'd be very appreciate that.

0

I resolve the problem using cvtos(ftocv(i, 0, 0));

thnaks you.

0

and  all of the ir matrices have the same dimensions.

please let me know how I convert my code more elegan.

0

Yes, GAUSS 10 does not have the function ntos. For older versions of GAUSS replace:

tmp = varget("vd"$+ntos(i));

with:

tmp = varget("vd"$+cvtos(ftocv(i, 1, 0)));

You must login to post answers.

Have a Specific Question?

Get a real answer from a real person

Need Support?

Get help from our friendly experts.

Try GAUSS for 14 days for FREE

See what GAUSS can do for your data

© Aptech Systems, Inc. All rights reserved.

Privacy Policy