How can I make this code more efficient?

close; clear all; clc;

global cp1 vp1 cp2 vp2 value gk;

A=1; % aggregate productivity
a=0.36; % capital share of output (alpha)
b=0.96; % subjective discounting (beta)
d=0.069;% depreciation


klow=0.1;
khigh=6;
knum=250;
kgrid = linspace(klow,khigh,knum);

uf = @(x)log(x);
pf = @(x)(A*(x.^a));

cp2=[cp2,pf(kgrid(1:knum))+((1-d)*kgrid(1:knum))]; %period 2
vp2=[vp2,uf(cp2(1:knum))];

for i=1:knum %period 1
   for j=1:knum
     cp1(i,j)=max(pf(kgrid(i))+(1-d)*kgrid(i)-kgrid(j),0);
     vp1(i,j)= log(cp1(i,j))+(b*vp2(j));
    end
  value=[value,max(vp1(i,:))];
  [val,loc] = max(vp1(i,:));
  gk=[gk,kgrid(loc)];
end

figure

subplot(211)
plot(kgrid, value)
hold on
title ( ' the value function ' )

subplot(212)
plot(kgrid, kgrid)
hold on
plot(kgrid, gk, '*')
title ( ' the decision rule ' )

saveas(gcf,'optimal2.png')

What can I do to make this code take up less lines or compute faster?

[jahanzaib ahmad]

no one gona help this way .please write what this code is about . edit ur title

[Stephen23]

"What can I do to make this code take up less lines or compute faster?"

• Write functions instead of scripts.
• Get rid of pointless cargo-cult programming close, clear all, and clc.
• Get rid of global variables and pass data properly as input/output arguments.
• Preallocate the output array/s before the loops, do not grow them inside the loops.
• Read and understand the techniques explained in the documentation:

https://www.mathworks.com/help/matlab/matlab_prog/techniques-for-improving-performance.html

• Use the profiler to identify which lines take the most time, then get back to us:

https://www.mathworks.com/help/matlab/matlab_prog/profiling-for-improving-performance.html