I want a correct logic of my code.Also, I want the operations on my two plots i.e. 1.)error of two plots 2.)difference curve 3.) correlation.

I want a correct logic of my code.

Also, I want the operations on my two plots i.e.

1. error of two plots
2. difference curve
3. correlation.

true=multibandread('paris.lan',[512,512,7],'uint8=>uint8',128,...
     'bil','ieee-le');

i1=true( :, :,1);
i2=true( :, :,2);
i3=true( :, :,3);
i4=true( :, :,4);
i5=true( :, :,5);
i6=true( :, :,6);
i7=true( :, :,7);

x(1)=double(i1(38,333));
x(2)=double(i1(38,334));
x(3)=double(i1(38,333));
x(1)=double(i1(38,333));
x(2)=double(i1(38,334));
x(3)=double(i1(39,333));
x(4)=double(i1(39,334));
y(1)=double(i1(486,115));
y(2)=double(i1(486,116));
y(3)=double(i1(487,115));
y(4)=double(i1(487,116));
avg1=((x(1)+x(2)+x(3)+x(4))/4);
avg2=((y(1)+y(2)+y(3)+y(4))/4);

   a(1)=double(i2(38,333));
  a(2)=double(i2(38,334));
  a(3)=double(i2(39,333));
  a(4)=double(i2(39,334));
  b(1)=double(i2(486,115));
  b(2)=double(i2(486,116));
  b(3)=double(i2(487,115));
  b(4)=double(i2(487,116));
  avg3=((a(1)+a(2)+a(3)+a(4))/4);
  avg4=((b(1)+b(2)+b(3)+b(4))/4);

m(1)=double(i3(38,333));
m(2)=double(i3(38,334));
m(3)=double(i3(39,333));
m(4)=double(i3(39,334));
n(1)=double(i3(486,115));
n(2)=double(i3(486,116));
n(3)=double(i3(487,115));
n(4)=double(i3(487,116));
avg5=((m(1)+m(2)+m(3)+m(4))/4);
avg6=((n(1)+n(2)+n(3)+n(4))/4);

p(1)=double(i4(38,333));
p(2)=double(i4(38,334));
p(3)=double(i4(39,333));
p(4)=double(i4(39,334));
q(1)=double(i4(486,115));
q(2)=double(i4(486,116));
q(3)=double(i4(487,115));
q(4)=double(i4(487,116));
avg7=((p(1)+p(2)+p(3)+p(4))/4);
avg8=((q(1)+q(2)+q(3)+q(4))/4);

r(1)=double(i5(38,333));
r(2)=double(i5(38,334));
r(3)=double(i5(39,333));
r(4)=double(i5(39,334));
s(1)=double(i5(486,115));
s(2)=double(i5(486,116));
s(3)=double(i5(487,115));
s(4)=double(i5(487,116));
avg10=((s(1)+s(2)+s(3)+s(4))/4);
avg9=((r(1)+r(2)+r(3)+r(4))/4);

h(1)=double(i6(38,333));
h(2)=double(i6(38,334));
h(3)=double(i6(39,333));
h(4)=double(i6(39,334));
j(1)=double(i6(486,115));
j(2)=double(i6(486,116));
j(3)=double(i6(487,115));
j(4)=double(i6(487,116));
avg12=((j(1)+j(2)+j(3)+j(4))/4);
avg11=((h(1)+h(2)+h(3)+h(4))/4);

f(1)=double(i7(38,333));
f(2)=double(i7(38,334));
f(3)=double(i7(39,333));
f(4)=double(i7(39,334));
g(1)=double(i7(486,115));
g(2)=double(i7(486,116));
g(3)=double(i7(487,115));
g(4)=double(i7(487,116));
avg13=((f(1)+f(2)+f(3)+f(4))/4);
avg14=((g(1)+g(2)+g(3)+g(4))/4);

%  d=[300,400,500];
% r=[avg1,avg3,avg5];
% plot(d,r)
% hold on
% c=[avg2,avg4,avg6];
% plot(d,c)
% plot(d,c)
% d=[100,200,300,400,500,600,700];
% r=[avg1,avg3,avg5,avg7,avg9,avg11,avg13];
% plot(d,r)
% hold on
% c=[avg2,avg4,avg6,avg8,avg10,avg12,avg14];
% plot(d,c)
d=[100,200,300,400,500,600,700];
r=[avg1,avg3,avg5,avg7,avg9,avg11,avg13];
% plot(d,r)
%hold on
c=[avg2,avg4,avg6,avg8,avg10,avg12,avg14];
% plot(d,c)
figure, plot(d,r,'r',d,c,'g')
title('actual signals');

z=corr2(r,c);
figure, plot(d,z)
title('correlation of 2 signals');   %correlation

mm=r-c;
SSE=dot(mm,mm);
figure, plot(d,mm)
title('error of 2 signals');    % error

deltaSignal = double(abs(c - r));
percentageDifference = double(deltaSignal ./ r); % Percent by element.
meanPctDiff = mean(percentageDifference); % Average percentage over all elements.
figure, plot(deltaSignal,percentageDifference,'r',deltaSignal,meanPctDiff,'g')
title('% difference of 2 signals');

[per isakson]

Is true the name of a variable?

[deepika]

yes,true is a variable name. Actually I just want to convert this long code of lines into a sjort one. and to make operations on the two plots obtained.

[deepika]

I tried but errors..

true=multibandread('paris.lan',[512,512,7],'uint8=>uint8',128,'bil','ieee-le');
for i=1:7
    for k=1:7
        true_value(i)=true(:,:,i);
        x(k)=true(:,:,i)(38,333);
        y(k)=true(:,:,i)(38,334);
        z(k)=true(:,:,i)(39,333);
        b(k)=true(:,:,i)(39,334);
        x1(k)=true(:,:,i)(486,115);
        y1(k)=true(:,:,i)(486,116);
        z1(k)=true(:,:,i)(487,115);
        b1(k)=true(:,:,i)(487,116);
    end
end
for a=1:7
    avg(a)=(x(a)+y(a)+z(a)+b(a))/4;
    avg1(a)=(x1(a)+y1(a)+z1(a)+b1(a))/4;
end

d=[1:1:10];
figure
plot( d, avg(a), 'r', d, avg1(a), 'g' )

[per isakson]

In Matlab, true is the name of a function. Try to avoid using names of Matlab functions as names of variables (or your own functions). Matlab lets you do it, but it may cause bugs, which are difficult to spot.

I modified your code so that it doesn't throw errors. However, it doesn't produce the expected result.

true=multibandread('paris.lan',[512,512,7],'uint8=>uint8',128,'bil','ieee-le');
for i=1:7
    for k=1:7
        true_value=true(:,:,i);
        x(k)=true_value(38,333);
        y(k)=true_value(38,334);
        z(k)=true_value(39,333);
        b(k)=true_value(39,334);
        x1(k)=true_value(486,115);
        y1(k)=true_value(486,116);
        z1(k)=true_value(487,115);
        b1(k)=true_value(487,116);
    end
end
for a=1:7
    avg(a)=(x(a)+y(a)+z(a)+b(a))/4;
    avg1(a)=(x1(a)+y1(a)+z1(a)+b1(a))/4;
end
%
d=[1:1:7];
figure
plot( d, avg, 'r', d, avg1, 'g' )

• Preallocate the variables, which now grows inside the loops. See Preallocation
• The rhs-expressions of the for k=1:7, ..., end doesn't depend on k.
• and more ...

[deepika]

Exactly,I am also not getting the expected output of this code.Help me a s soon as possible.

[deepika]

Please let me know about the 3 operations,which I asked above.

[per isakson]

"I tried but errors." I tried to help you in that effort.

However, I'm not clear about your three items and I'm not prepared to analyze the 100+ lines of code in your question.

[deepika]

3 operations are

1. to calculate error of the two plots
2. to calculate difference of the two plots
3. to calculate correlation of the two plots

[Aarti Dwivedi]

Why can't you get the difference by just subtracting the variables from each other? I am assuming they are of the same size

[deepika]

Thank you so much.