error synthax : "else", "startsp" 'endif'

It appears that the problem is that many of the statements in this program do not end with semi-colons. In GAUSS, statements are separated by semi-colons. For example, towards the top of this program, we see:

if freq == 1
   fsp = sp
   fep = ep
   ffobs = fobs
else

Each of these lines should be separated by semi-colons, otherwise GAUSS interprets them as one long statement, like this:

if freq == 1 fsp = sp fep = ep ffobs = fobs else

The line above is not a valid GAUSS statement. To make it valid, GAUSS code, we need to add semi-colons to the end of the lines. Here is a fixed version of the first bit of the file.

new 2000,320000;

/*  Program version: May 31, 1999  */

run C:\gauss10\Balcilar\msvar\setup.mod;

/* Fix data matrix and obs to match selected sample  */

load data[434,2]=C:\gauss10\Balcilar\msvar\data\ymm5995.asc;

if freq == 1;
   fsp = sp;
   fep = ep;
   ffobs = fobs;
else;
   fsp = sp-(1/(2*freq));
   fep = ep-(1/(2*freq));
   ffobs = fobs-(1/(2*freq));
endif;

Dear Jason Many Thanks,

I added to the code the sem-clons but I have got another error message:

Line 83 in C:\gauss10\Balcilar\msvar.txt
Illegal redefinition of procedure G0156 : 'm'
Line 86 in C:\gauss10\Balcilar\msvar.txt
Illegal redefinition of procedure G0156 : 'm'
Line 89 in C:\gauss10\Balcilar\msvar.txt
Illegal redefinition of procedure G0156 : 'm'

Thanks a lot for your help

 
*******************************************************************************************
 
 
 
new 2000,320000
 
/*  Program version: May 31, 1999  */
 
run C:\gauss10\Balcilar\msvar\setup.mod
 
/* Fix data matrix and obs to match selected sample  */
 
load data[434,2]=C:\gauss10\Balcilar\msvar\data\ymm5995.asc;
 
if freq == 1;
fsp = sp;
fep = ep;
ffobs = fobs;
else;
fsp = sp-(1/(2*freq));
fep = ep-(1/(2*freq));
ffobs = fobs-(1/(2*freq));
endif;
 
startsp = round((((fsp-ffobs)*freq)+1));
endsp = round((((fep-ffobs)*freq)+1));
data = data[startsp:endsp,.];
obs = rows(data);
 
/* Prefilter seasonals from the data using dummies   */
 
if preremov == 1;
run C:\gauss10\Balcilar\msvar\programs\searemov.pgm;
endif;
 
if m == 1 or m == 2 or m == 3 or m == 4 or m == 5;
m1 = 1;
m2 = 0;
k = q^2;
k1 = q;
k2 = 1;
elseif m == 6;
m1 = 1;
m2 = 0;
k = q^(p+1);
k1 = q^p;
k2 = q^(p-1);
elseif m == 7 or m == 8 or m == 9;
m1 = 1;
m2 = 1;
k = q^2;
k1 = q;
k2 = 1;
else;
m1 = 0;
m2 = 0;
endif;
if m1 == 0;
print " ";
print "You have NOT selected a model. Must be Q playing with your mind again!";
end;
endif;
if m2 == 0 and p == 0;
print " ";
print "You have forgotten to select lags. Must be all that Romulanale last night!";
end;
endif;
if dv == 0 and Sea ne 0;
print "You have selected dummy seasonals and NO deterministic variables(dv=0).Program NOT executed!";
end;
elseif dv == 0 and Trend ne 0;
print " ";
print "You have selected a linear deterministic trend and NO deterministic variables (dv=0).Program NOT executed!";
end;
endif;
if qs >= 2 and q ne 2^qs;
print " ";
print "Your specifications of q and qs are NOT compatible.Programs NOT executed!";
end;
endif;
if qs >= 2 and sump == 1;
print "You have selected  sump=1  and qs>1. Option not available.Program NOT executed!";
end;
endif;
if m == 7;
mh = 7;
m = 1;
elseif m == 8;
mh = 8;
m = 2;
elseif m == 9;
mh = 9;
m = 5;
endif;
 
if Trend == 0 and Sea == 0;
trend1 = 1;
else;
trend1 = 0;
endif;
 
if trend1 == 1 and dv == 1;
print "You have selected deterministic variables (dv=1), but not specified a linear trend or seasonals.Program NOT executed!";
end;
endif;
 
/*   Constructing data matrices   */
 
T = obs-p;
 
if dv == 1 and sea == 0;
d = Trend;
elseif dv == 1 and sea == 4 or sea == 12;
d = Trend+sea-1;
else;
d = 0;
endif;
 
y = data[(p+1):obs,.]';
if p >= 1;
xmat = data[p:(obs-1),.]';
j = 2;
do while j <= p;
xmat = xmat|(data[(p+1-j):(obs-j),.]');
j = j+1;
endo;
endif;
 
conmat = ones(1,T);
 
/*  Defining the variable dtrend */
 
if Trend == 1 and d >= 1;
dtrend = seqa(1,1,T)';
/* we read the data for conditioning variable */
dtrend = zmat;
d = cols(zmat);
endif;
 
/*  Defining the centered seasonal dummies  */
 
if sea == 4 and d >= 3;
seamat = zeros(3,T);
j = 1;
do while j <= 3;
i = j;
do while i <= T;
seamat[j,i] = 1;
i = i+4;
endo;
j = j+1;
endo;
adjsea = (1/4)*ones(3,T);
seamat = seamat - adjsea;
elseif sea == 12 and d>= 11;
seamat = zeros(11,T);
j = 1;
do while j <= 11;
i = j;
do while i <= T;
seamat[j,i] = 1;
i = i+12;
endo;
j = j+1;
endo;
adjsea = (1/12)*ones(11,T);
seamat = seamat - adjsea;
endif;
 
if Trend == 1 and sea == 4 or sea == 12;
dmat = dtrend|seamat;
elseif Trend == 0 and sea == 4 or sea == 12;
dmat = seamat;
elseif Trend == 1 and d == 1;
dmat = dtrend;
endif;
 
if d == 0 and p >= 1;
ymat = conmat|xmat;
elseif d == 0 and p == 0;
ymat = conmat;
elseif d >= 1 and p >= 1;
ymat = conmat|dmat|xmat;
else;
ymat = conmat|dmat;
endif;
 
if m >= 3 and m <= 4 and d >= 1;
xmat = dmat|xmat;
endif;
 
/* Setting up the n*n by n*(n+1)/2 Duplication matrix  */
 
Dupn = zeros(n*n,n*(n+1)/2);
dnel2 = 0;
j = 1;
do while j <= n;
i = 1;
do while i <= n-j+1;
dnel2 = dnel2+1;
Dupn[i+(n+1)*(j-1),dnel2] = 1;
i = i+1;
endo;
j = j+1;
endo;
if n > 1;
j = 1;
do while j <= n-1;
dnstar = 1;
i = 1;
do while i <= j;
Dupn[i+n*j,dnstar+j-i+1] = 1;
dnstar = dnstar+n-i+1;
i = i+1;
endo;
j = j+1;
endo;
endif;
DMP = invpd(Dupn'*Dupn)*Dupn';
 
/*  Setting up a selection matrix for mu(i) in model 6  */
 
if m == 6;
musel = eye(q) .*. ones(1,q^p);
j = 1;
do while j <= p;
muselj = ones(1,q^j) .*. (eye(q) .*. ones(1,q^(p-j)));
musel = musel|muselj;
j = j+1;
endo;
commute = zeros(q*(p+1),q*(p+1));
i = 1;
do while i <= p+1;
j = 1;
do while j <= q;
commute[j+q*(i-1),i+(p+1)*(j-1)] = 1;
j = j+1;
endo;
i = i+1;
endo;
musel = commute'*musel;
endif;
 
/*  Starting the iterations   */
 
if rest >= 1;
lambda = zeros(rest,1);
if m == 1;
selV = eye(n)~zeros(n,n*(q-1));
theta = vec(selV*A);
omega = DMP*vec(selV*V);
i = 2;
do while i <= q;
selVi = zeros(n,n*q);
j = 1;
do while j <= n;
selVi[j,j+n*(i-1)] = 1;
j = j+1;
endo;
theta = theta|(vec(selVi*A));
omega = omega|(DMP*vec(selVi*V));
i = i+1;
endo;
theta = theta|omega;
elseif m == 2;
selV = eye(n)~zeros(n,n*(q-1));
theta = vec(selV*A);
omega = DMP*vec(selV*V);
i = 2;
do while i <= q;
selAi = zeros(n,n*q);
j = 1;
do while j <= n;
selAi[j,j+n*(i-1)] = 1;
j = j+1;
endo;
theta = theta|(vec(selAi*A));
i = i+1;
endo;
theta = theta|omega;
elseif m == 3;
mu = A[.,1];
selAr = eye(n)~zeros(n,n*(q-1));
selAc = zeros(1,n*p+d)|eye(n*p+d);
Amat = selAr*A*selAc;
theta = mu|(vec(Amat));
i = 1;
do while i <= q;
selVi = zeros(n,n*q);
j = 1;
do while j <= n;
selVi[j,j+n*(i-1)] = 1;
j = j+1;
endo;
Vmati = selVi*V;
theta = theta|(DMP*vec(Vmati));
i = i+1;
endo;
elseif m == 4;
mu = A[.,1];
selAr = eye(n)~zeros(n,n*(q-1));
selAc = zeros(1,n*p+d)|eye(n*p+d);
Amat = selAr*A*selAc;
Vmat = selAr*V;
theta = mu|(vec(Amat))|(DMP*vec(Vmat));
elseif m == 5;
selV = eye(n)~zeros(n,n*(q-1));
theta = vec(selV*A);
omega = DMP*vec(selV*V);
i = 2;
do while i <= q;
selVi = zeros(n,n*q);
j = 1;
do while j <= n;
selVi[j,j+n*(i-1)] = 1;
j = j+1;
endo;
omega = omega|(DMP*vec(selVi*V));
i = i+1;
endo;
theta = theta|omega;
elseif m == 6;
mu = A[.,1];
selV = eye(n)~zeros(n,n*(q-1));
selAc = zeros(1,n*p+d)|eye(n*p+d);
Amat = selV*A*selAc;
Vmat = selV*V;
theta = mu|(vec(Amat))|(DMP*vec(Vmat));
endif;
fthetab = Rmat*theta-rvec;
restchek = maxc(abs(fthetab));
if restchek ne 0 and check == 1;
print " "
print "The initial parameter values you have selected do NOT meet the restrictions. Program NOT executed!";
end;
else;
Rort = null(Rmat);
endif;
endif;
 
Ainit = A;
Vinit = V;
Pinit = Pmat;
rho = (1/k)*ones(k,1);
lnL = 0;
dlnL = 0;
maxpar = 100;
maxgr = 100;
 
begini = sp;
endini = sp+((p-1)/freq);
begest = sp+(p/freq);
endest = ep;
begini1 = trunc(begini);
endini1 = trunc(endini);
begest1 = trunc(begest);
endest1 = trunc(endest);
if freq ne 1;
begini2 = begini+(1/(2*freq));
endini2 = endini+(1/(2*freq));
begest2 = begest+(1/(2*freq));
endest2 = endest+(1/(2*freq));
begini3 = round(((begini2-begini1)*freq));
endini3 = round(((endini2-endini1)*freq));
begest3 = round(((begest2-begest1)*freq));
endest3 = round(((endest2-endest1)*freq));
estsamp = "Estimation period:                     "$+ ftocv(begest1,2,0)$+ ":"$+ ftocv(begest3,1,0)$+ "-"$+ ftocv(endest1,2,0)$+ ":"$+ ftocv(endest3,1,0);
if p == 0;
inisamp = "Initial observations:                  None";
else;
inisamp = "Initial observations:                  "$+ ftocv(begini1,2,0)$+ ":"$+ ftocv(begini3,1,0)$+ "-"$+ ftocv(endini1,2,0)$+ ":"$+ ftocv(endini3,1,0);
endif;
else;
estsamp = "Estimation period:                     "$+ ftocv(begest1,2,0)$+ "-"$+ ftocv(endest1,2,0);
if p == 0;
inisamp = "Initial observations:                  None";
else;
inisamp = "Initial observations:                  "$+ ftocv(begini,2,0)$+ "-"$+ ftocv(endini,2,0);
endif;
endif;
mstr = "Variables for model "$+ ftocv(m,1,0)$+ ":";
estsamp1 = "          "$+ estsamp;
inisamp1 = "          "$+ inisamp;
numobs = "Number of observations:                "$+ ftocv(T,2,0);
numobs1 = "          "$+ numobs;
 
nstates1 = seqa(1,1,q);
if qs >= 2;
statesel = eye(2) .*. ones(1,2^(qs-1));
i = 1;
do while i <= qs-1;
statesi = ones(1,2^i) .*. (eye(2) .*. ones(1,2^(qs-1-i)));
statesel = statesel|statesi;
i = i+1;
endo;
comsel = zeros(2*qs,2*qs);
i = 1;
do while i <= qs;
j = 1;
do while j <= 2;
comsel[j+2*(i-1),i+qs*(j-1)] = 1;
j = j+1;
endo;
i = i+1;
endo;
statesel = comsel'*statesel;
statesej = reshape(statesel[.,1],2,qs)';
state1 = seqa(1,1,2);
states = statesej*state1;
j = 2;
do while j <= q;
statesej = reshape(statesel[.,j],2,qs)';
states = states~(statesej*state1);
j = j+1;
endo;
nstates1 = nstates1~(states');
endif;
 
if qs == 1;
ejstr = "             State         Est #obs";
elseif qs >= 2;
ejstr = "             State       Regime var's 1 to "$+ ftocv(qs,1,0)$+ "        Est #obs";
endif;
 
headmsg = "       ESTIMATING MS("$+ ftocv(q,1,0)$+ ")-VAR("$+ ftocv(p,1,0)$+ ") MODEL "$+ ftocv(m,1,0)$+ " FOR: "$+ mvars;
headund = "       =";
headw = strlen(headmsg);
i = 9;
do while i <= headw;
headund = headund$+ "=";
i = i+1;
endo;
 
output off;
format /M1 /ROS 10,6;
 
cls;
locate csrlin,csrcol;
print ""
print headmsg;
print headund;
print " "
 
it = 1;
iterate: do while it<=maxit and maxpar>convcrit and maxgr>convcrit and m1==1;
 
run C:\gauss10\Balcilar\msvar\programs\smooth.pgm;
run C:\gauss10\Balcilar\msvar\programs\gradient.pgm;
if maxgr > convcrit;
run C:\gauss10\Balcilar\msvar\programs\estimate.pgm;
endif;
if it == 1;
values = "      "$+ ftocv(it,1,0)$+ "   "$+ ftocv(lnL,1,6)$+ "   "$+ ftocv(dlnL,1,6)$+ "   "$+ ftocv(maxpar,1,6)$+ "   "$+ ftocv(maxgr,1,6);
elseif it < 10 and it > 1;
values1 = "      "$+ ftocv(it,1,0)$+ "   "$+ ftocv(lnL,1,6)$+ "   "$+ ftocv(dlnL,1,6)$+ "   "$+ ftocv(maxpar,1,6)$+ "   "$+ ftocv(maxgr,1,6);
values = values$| values1;
elseif it < 100 and it > 9;
values1 = "     "$+ ftocv(it,1,0)$+ "   "$+ ftocv(lnL,1,6)$+ "   "$+ ftocv(dlnL,1,6)$+ "   "$+ ftocv(maxpar,1,6)$+ "   "$+ ftocv(maxgr,1,6);
values = values$| values1;
else;
values1 = "    "$+ ftocv(it,1,0)$+ "   "$+ ftocv(lnL,1,6)$+ "   "$+ ftocv(dlnL,1,6)$+ "   "$+ ftocv(maxpar,1,6)$+ "   "$+ ftocv(maxgr,1,6);
values = values$| values1;
endif;
 
nstates = nstates1~(piT*imathT);
nstates2 = "          "$+ ftocv(nstates,1,0);
 
itmsg = "          Iteration number:                      "$+ ftocv(it,1,0);
lnlmsg = "          Current value of log Likelihood:       "$+ ftocv(lnL,1,6);
mparmsg = "          Maximum absolute parameter change:     "$+ ftocv(maxpar,1,8);
mgrmsg = "          Maximum absolute gradient element:     "$+ ftocv(maxgr,1,8);
print itmsg;
print " "
print lnlmsg;
print mparmsg;
print mgrmsg;
if p >= 1;
karlmsg = "          Largest system eigenvalue:             "$+ ftocv(maxc(Amod),1,6);
karlmsg2 = "          (Stationarity condition)";
print karlmsg;
print karlmsg2;
endif;
print " "
print estsamp1;
print inisamp1;
print numobs1;
print " ";
print ejstr;
print nstates2;
print " "
if p >= 1;
backline=csrlin-15-q;
else;
backline=csrlin-13-q;
endif;
locate backline,1;
it = it+1;
endo;
 
output on;
screen on;
print "A =" A;
print "V =" V;
 
if m1 == 1 and it <= maxit;
run C:\gauss10\Balcilar\msvar\programs\conscore.pgm;
run C:\gauss10\Balcilar\msvar\programs\standerr.pgm;
if rest == 0 and sump ne 1;
run C:\gauss10\Balcilar\msvar\programs\misspec.pgm;
endif;
infocrit = (2*((s-rest)-lnL))/T;
bic = (2*(((s-rest)*ln(T))-lnL))/T;
lil = (2*(((s-rest)*ln(ln(T)))-lnL))/T;
infocrit = infocrit~bic~lil;
endif;
output on;
screen off;
if m2 == 1;
m = mh;
endif;
if m1 == 1 and it > maxit;
screen on;
output off;
if p>=1;
forward=csrlin+15;
else;
forward=csrlin+13;
endif;
locate forward,1;
print ""
print ""
print "          ***************************************************************";
print "          ***************************************************************";
print "          ***                                                         ***";
print "          ***   W A R N I N G !   W A R N I N G !   W A R N I N G !   ***";
print "          ***                                                         ***";
print "          ***************************************************************";
print "          ***************************************************************"
print ""
print ""
print "                      **** ITERATIONS HAVE NOT CONVERGED ****";
print " "
print ""
print "If you would like to continue the iterations, please enter a new value for";
print "the maximum number of iterations (otherwise, select any integer and press"
print "enter)."
print ""
maxitnew = con(1,1);
maxitnew = trunc(maxitnew);
if maxitnew > maxit;
maxit = maxitnew;
 
cls;
locate csrlin,csrcol;
print ""
print headmsg;
print headund;
print " "
 
goto iterate;
else;
screen off;
output on;
endif;
 
/* Perhaps add some options */
 
endif;
 
if m1 == 1;
print iteration   lnL    difflnL  maxpar diff  maxgradient;
print values;
endif;
if m1 == 1 and it <= maxit;
print " "
print "********************************************************************"
print " "
print mstr;
print " "
print mvars;
print " "
print estsamp;
print inisamp;
print " "
print numobs;
print " "
print ejstr;
print nstates2;
print " "
if p >= 1;
if (maxc(Amod)) >= 1;
print "WARNING!       The estimated parameters are not consistent with the 2nd order stationarity condition of Karlsen.";
print " "
endif;
if (maxc(Amod)) < 1 and d == 0;
run C:\gauss10\Balcilar\msvar\programs\meanvar.pgm;
run C:\gauss10\Balcilar\msvar\programs\mvstd.pgm;
endif;
endif;
print "Initial values of A";
print Ainit;
print " "
print "Initial values of V";
print Vinit;
print " "
print "Initial values of Pmat";
print Pinit;
print " "
print "These are the ML estimates of A";
print A;
print " "
print "Standard errors based on the conditional scores";
print Aerror1;
print " "
print "These are the ML estimates of V";
print V;
print " "
print "Standard errors based on the conditional scores";
print Verror1;
print " "
print "These are the ML estimates of the Markov probabilities";
if qs == 1;
print Pmat;
print " "
else;
print p0q;
print " "
endif;
print "Standard errors based on the conditional scores";
print Perror1;
print " "
if rest == 0 and p>=1;
print "Nyblom tests for parameter stability";
print " "
print "   Markov       .90        .95         .99";
print Np0;
print " "
print "      mu        .90        .95         .99";
print Nmu;
print " "
print "   Res cov      .90        .95         .99";
print Noma;
print " "
print "Nyblom tests for parameter stab in eq (rows) / state (cols)";
print Nyeqst;
print " "
if n*p+d+2 <= 20;
print "              Critical values";
print "      .90        .95        .99";
print Nyeqstcv;
print " "
endif;
endif;
if sump ne 1;
print "Wald test of serially uncorrelated Markov chain";
print " "
print Wald;
print " "
print "with marginal level of significance";
print " "
print Waldpv;
print " "
print "F-test of serially uncorrelated Markov chain";
print " "
print Waldf;
print " "
print "with marginal level of significance";
print " "
print Waldfpv;
print " "
endif;
if qs >= 2;
print "Wald and F-tests of no serial correlation for each independent Markov chain";
print " "
print "regime variable      Wald            F";
print (seqa(1,1,qs))~Wald1~Wald1f;
print " "
print "with marginal levels of significance";
print " "
print (seqa(1,1,qs))~Wald1pv~Wald1fpv;
print " "
endif;
if p >= 1;
if (maxc(Amod)) < 1 and d == 0;
print "First two moments conditional on s(t)";
print "     Mean         Covariance";
print dely~convy;
print " "
print "Standard errors based on conditional scores";
print "and the Delta method using numerical partials";
print stddely~seconvy;
print " "
print "First two moments";
print "     Mean         Covariance";
print muy~vy;
print " "
print "Standard errors based on conditional scores";
print "and the Delta method using numerical partials";
print stdmuy~stdvy;
print " "
if sump ne 1;
print "Ergodic probabilities";
print " "
print pivec';
print " "
print "Standard errors based on conditional scores";
print " "
print pierr1;
print " "
endif;
endif;
print "Modulus of the eigenvalues of the dynamic model";
print "(largest should be less than 1 for stationarity)";
print Amod;
print " "
endif;
print "       Information Criteria";
print "     AIC        BIC        LIL";
print " "
print infocrit;
print " "
if qs >= 2;
lms1 = "The LM test for "$+ ftocv(qs,1,0)$+ " independent";
print lms1;
print "two-state regime variables";
print " "
print LM;
print " "
print "with marginal level of significance";
lms2 = "(for "$+ ftocv(s11-s1,1,0)$+ " degrees of freedom)";
print lms2;
print " "
print LMpv;
print " "
print "The small sample corrected F-statistic";
print " "
print LMf;
print " "
print "with marginal level of significance";
print " "
print LMfpv;
print " "
endif;
if rest >= 1 or sump == 1;
if rest >= 1 and sump == 1;
lmr1 = "The LM test of "$+ ftocv(rest,1,0)$+ " restrictions on the VAR parameters and "$+ ftocv(s1*s1,1,0)$+ " rest-restrictions on the Markov probabilities";
elseif rest >= 1 and sump ne 1;
lmr1 = "The LM test of "$+ ftocv(rest,1,0)$+ " restrictions on the VAR parameters";
elseif rest == 0 and sump == 1;
lmr1 = "The LM test of a serially uncorrelated Markov process ("$+ ftocv(s1*s1,1,0)$+ " restrictions on P)";
endif;
print lmr1;
print " "
print LMrest;
print " "
print "with marginal level of significance";
print " "
print LMrestpv;
print " "
print "The small sample corrected F-statistic";
print " "
print LMfres;
print " "
print "with marginal level of significance";
print " "
print LMfrespv;
print " "
endif;
if rest == 0 and sump ne 1;
print "Misspecification tests based on conditional scores";
print " "
if qs == 1;
print "         Equation-by-equation tests";
print "  Equation   Autocorr     ARCH      Markov";
print misspece;
print " "
print "               with p-values";
print pvspece;
print " "
if n >= 2;
print "           System tests";
print "  Autocorr     ARCH       Markov";
print " "
print misspec;
print " "
print "          with p-values";
print " "
print pvspec;
print " "
endif;
elseif qs >= 2;
if n == 1;
print "              Equation-by-equation tests";
print "    Equation     Autocorr         ARCH       Markov";
print misspece;
print " "
print "                    with p-values";
print pvspece;
print " "
elseif n >= 2;
print "        Equation-by-equation tests";
print "    Equation     Autocorr         ARCH";
print misspece;
print " "
print "              with p-values";
print pvspece;
print " "
print "                System tests";
print "    Autocorr         ARCH       Markov";
print " "
print misspec;
print " "
print "               with p-values";
print " "
print pvspec;
print " "
endif;
endif;
endif;
endif;
screen on;
output off;
format /MB1 /ROS 16,8;
if m1 == 1 and it <= maxit;
run C:\gauss10\Balcilar\msvar\programs\graphs.pgm;
screen off;
output on;
format /M1 /ROS 10,6;
run C:\gauss10\Balcilar\msvar\programs\regest.pgm;
output off;
screen on;
endif;
system;