Help rewriting a function using If statements and For Loop

Hello everyone,

So I have the following code, which is a simple Quantization, that check the value of each element in an array, and round it to an already set value, here I have only 4 levels (1.5 0.5 -0.5 -1.5). And I am assuming that the max amplitude of the input signal is 2.

dq_tx = [];

for i1 = 1:numel(k)

if (d_k(i1) >=1)

dqe = 1.5;

elseif (d_k(i1) >=0)

dqe = 0.5;

elseif (d_k(i1) >=-1)

dqe = -0.5;

else (d_k(i1) <-1);

dqe = -1.5;

end

dq_tx = [dq_tx dqe];

end

What I want to do is making it more global, to have the number of levels as an input from the user, and make it depended on the max ampltiude of the input signal as well. And here is what I have so far in terms of the conceptual implicaiton, but It's not working yet and I feel i am missing something

amp=1.2; %max amplitude of the signal as input from the user

L = 4; % # levels in the quantizer

dMin = -amp;

dMax = amp;

%q = 2*dMax/L; % step size

%q2 = q/2;

dR = linspace(dMin,dMax,L+1) % decision intervals

dR2=zeros(1,L);

i=0;

for i=1:L

dR2(i) = ( dR(i) + dR(i+1)) ./ 2; %Re-adjusting the intervals values

end

d_k=[1 0.5 -0.2 -0.7 -1] %Assumed message signal values

%Qunatization Code

dq_tx = [];

for i1 = 1:numel(k)

if (d_k(i1) >=dR2(end))

dqe = dR2(end);

elseif

for i2=1:L-2

if (d_k(i1) >=dR2(end-i2))

dqe = dR2(end-i2);

end

end

else (d_k(i1) <dR2(1)); % Correct the greater than sign

dqe = dR2(1);

end

dq_tx = [dq_tx dqe];

end

[Jan]

What about: [BINS, EDGES] = discretize(X,N) ?

[Stephen23]

"I feel i am missing something "

https://www.mathworks.com/help/matlab/ref/histcounts.html

[Abdelrhman Abdelfatah]

thank you so much, that was very useful. I tried to use it but the issue I faced is later on I plan to use either a midtread or midrise quantizer based on the user's perfernce, and I would need to have a direct role in deciding the steps, and it actually shouldn't be equally split, but at least applicable, especialyl that I would need to do decoding later on.

[Abdelrhman Abdelfatah]

Thank you so much, I figured it out and it's working like magic now!

[Stephen23]

"I tried to use it but the issue ... it actually shouldn't be equally split..."

Neither DISCRETIZE nor HISTCOUNTS require the edges to be "equally split". They are just simpler ways of doing what you want, without needing to reinvent the wheel.

Answer 1

[James Water Bird]

to make your code more global and adaptable to different input signal amplitudes and number of quantization levels, you can make the following changes to your code:

• Use a for loop to iterate over the values in the input signal d_k and apply the quantization rules to each value.
• Use the input signal amplitude amp and number of levels L to calculate the step size q and decision intervals dR.
• Use an if statement to compare the current input value d_k(i1) to the decision intervals dR and determine which quantized value dqe to use.

Forexample

% Input parameters

amp = 1.2; % max amplitude of the input signal

L = 4; % number of quantization levels

% Calculate step size and decision intervals

q = 2*amp/L; % step size

dR = linspace(-amp, amp, L+1); % decision intervals

% Input signal

d_k = [1 0.5 -0.2 -0.7 -1];

% Quantization

dq_tx = []; % initialize quantized output signal

for i1 = 1:numel(d_k)

dqe = 0; % initialize quantized value

% Compare input value to decision intervals

if d_k(i1) >= dR(end) % greater than or equal to maximum interval

dqe = dR(end); % use maximum interval value

else

for i2 = 1:L-1

if d_k(i1) >= dR(end-i2) && d_k(

[Abdelrhman Abdelfatah]

Okay, that's actually a better implementation, how can I go on with the rest of the code?

I went as, however iti keeps giving me 0 output values

for i2 = 1:L-1
            if (d_k(i1) >= dR(end-i2)) && (d_k(i1) <= dR(end-i2-1))
                dqe = dR(end-i2);
            end 
           dq_tx = [dq_tx dqe];
        end

[Abdelrhman Abdelfatah]

Thank you so much, I figured it out and it's working like magic now!