Misbehaving script

Hi, i'm writing a script to calculate diameters of pipes for me. However, Matlab is acting up on me (or i'm missing something).

I've set up a few parameters at the start of the script, defined a function and set up a for loop to get me my values.

Q_in=(2/9);
n_nozzle=10;
n_row=9;
d_in=250e-3;
Q_nozzle=2.2e-3;
Q=zeros(n_row,n_nozzle);
Q(1,1)=Q_in;

Q_R=@(r,c) Q(r-1,c)-(Q_nozzle+(n_nozzle*Q_nozzle))

for (i=2:n_row)
  Q(i,1)=Q_R(i,1);
end

for the first iteration where i=2, the script works properly and produces the right answer. But starting from i=3, the Q(r-1,c) part of the Q_R function just keeps ending up as 0 where it should be referencing Q(2,1).

I have manually tried Q(2,1)-(Q_nozzle+(n_nozzle*Q_nozzle)) and in this case the Q(2,1) correctly references Q(2,1). The function just does not work properly if I use Q_R(3,1).

What is going on here ?

Answer 1

[Teja Muppirala]

In your function handle, the "Q" that it is referencing is the original Q at the time you created the function handle. It is not linked with whatever is in your workspace.

The following should make this clear.

A = 10;

F = @(x) A + x
F(7)

A = 0;
F(7)

F = @(x) A + x
F(7)

Even though I change A, the function handle does not reflect that change until I redefine the function handle.

[Paulo Silva]

That was a tricky one, thanks for finding and explaining the error, +1 vote

Answer 2

[Paulo Silva]

I see it working properly, see it by yourself with this code

clc
Q_in=(2/9);
n_nozzle=10;
n_row=9;
d_in=250e-3;
Q_nozzle=2.2e-3;
Q=zeros(n_row,n_nozzle);
Q(1,1)=Q_in;
Q_R=@(r,c) Q(r-1,c)-(Q_nozzle+(n_nozzle*Q_nozzle));
for i=2:n_row
    Q(i,1)=Q_R(i,1);    
    i
    Q_R(i,1)
    pause(0.3)
end

[Christopher]

When I run the code Q(1,1) is 2/9 as I defined.

Q(2,1)=0.1980 as it should be

but Q(3,1)=-0.0242 which is wrong.

It should be doing 0.1980-0.0242 = 0.1738

But instead it is doing 0-0.0242=-0.0242

[Paulo Silva]

ok you are right and I'm stuck without a solution, it's been a while since I had this kinds of problems to solve but I won't give up

[Paulo Silva]

So it works diferently without the function

Q_in=(2/9);

n_nozzle=10;

n_row=9;

d_in=250e-3;

Q_nozzle=2.2e-3;

Q=zeros(n_row,n_nozzle);

Q(1,1)=Q_in;

for n=2:n_row

Q(n,1)=Q(n-1,1)-(Q_nozzle+(n_nozzle*Q_nozzle));

end

Q(1:end,1)

[Christopher]

Yeah, i've already changed my code to work just like you have done. But it is still puzzling me why it does not work with the function ?

[Paulo Silva]

so am I, been away from such weird problems for some time, you can reproduce the original problem without the loop

Q_in=(2/9);

n_nozzle=10;

n_row=9;

d_in=250e-3;

Q_nozzle=2.2e-3;

Q=zeros(n_row,n_nozzle);

Q(1,1)=Q_in;

%for

n=2:n_row;

Q(n,1)=Q(n-1,1)-(Q_nozzle+(n_nozzle*Q_nozzle));

%end

Q(1:end,1)

very strange or I might be missing something simple :S