convolute two functions(two signals)

hi everbody i'd like to make a script file which convolute to two function . so i started to code this but unfortunately it doesn't work can you help me thx in advanced

here is my source code.

function x = conv(t) 

% x = 1 pour t compris en -pi/2 et pi/2 et 0 ailleurs 

if (t<abs(pi/2))
x=1;
else
    x=0;
end

y=conv(x,x);
plot(y)

Answer 1

[taftoof]

function x = myconv(t) 

x = abs(t) < pi/2;
end

w=conv(x,x);
plot(w)

like that what're you think?

[Walter Roberson]

You should not have the "end" in the middle there.

[taftoof]

what do u suggest to change on code

[Image Analyst]

He meant to take out the "end", okay? Now, what do you want to use in your main program after calling myconv()? Do you want x, w, or both?

[taftoof]

what i'd like is to:

1 create a function x (for exemple f(x)=x)

after that convolute x with x

put the result in w and plot it

[Walter Roberson]

function x = myconv(t)

x = abs(t) < pi/2;

w=conv(x,x);

plot(w)

[Image Analyst]

You can "almost exactly" do what you say like this

function x(t) % Note: no output arguments will be passed out

small_x = abs(t) < pi/2; % Can't also call this x since function name is x.

w = conv(small_x, small_x);

plot(w);

Since, after I asked, you said you don't need any output argument and that you now want the function to be called "x" instead of "myconv." I say "almost" because you can't do exactly what you say because you said the function should be called x but then inside you want x to be convolved (not "covoluted") with x which would be a recursive mess, IF it could even be done at all. But I think Walter's code is better.

Answer 2

[Walter Roberson]

What are the two functions? You are convolving something with itself.

Is t a vector or a scalar? "if" applied to a vector is treated as if you had coded "if all(....)" and so is true only if the logical condition applies to all the logical vector values. As it probably does not, you are assigning the scalar 0 to x. Hint: read about logical indexing.

Answer 3

[Image Analyst]

Maybe you mean

x = t < pi/2;  % x will be a logical matrix.
y = conv(double(x),  double(x));

not really sure -- seems like a weird thing to want to do. You might also be interested in the xcorr() function.

Answer 4

[taftoof]

hi guys

here is my function

x(t)=1 if -pi/2<t<pi/2 x(t)=0 otherwise

and i'd like to convoluate x(t) with x(t) conv(x(t),x(t)) thx in advanced

[Walter Roberson]

Take Image Analyst's code and add abs() around the t.

[taftoof]

thx for ur help

it's done but matlab show me this message

Error using conv (line 7)

Not enough input arguments.

function x = conv(t)

% x = 1 pour t compris en -pi/2 et pi/2 et 0 ailleurs

if (t<(abs(pi/2)))

x=1;

else

x=0;

end

y=conv(x,x);

plot(y)

[Walter Roberson]

Do not call your own routine "conv" ! You are clashing with MATLAB's "conv" routine, and your routine is trying to call itself!

[taftoof]

how can i change that pls

[Image Analyst]

Not to mention the fact that your original code, my code, and your second code here don't do that you asked for. Do you know that abs(pi/2) is the same as pi/2? Do you know that t<abs(pi/2) is the same as t<pi/2? But it appears NOW that you have a different condition and that is "-pi/2<t<pi/2" I hope you can figure out why that is NOT THE SAME as t<abs(pi/2). So, once you've understood that you can do

x = abs(t) < pi/2;

That will set x = 1 even if t = -1, whereas your and my original code won't (it does your original criteria instead, which is different).

[Walter Roberson]

Change your "function x = conv(t)" line to "function x = myconv(t)" and save the result in the file myconv.m and then delete the conv.m that you created.