Not sure how to do a smoothing average of noisy data

Not sure how to do a smoothing average of noisy data based off what the instructions asked this is what I've come up with so far any help would be appreciated.

clear;
clc;

load('noisydata.mat','x');
N = x;
y = zeros(1,length(N));
for k = 2:1:length(x)
    if (k~= 1 & k~= N)
        y(k) = (x(k-1)+x(k+1))/2;
    end 
end

Answer 1

[Image Analyst]

You forgot to set y(1) = x(1) before the loop, and, after the loop, set y(length(x)) = x(end). Then have your loop go from k = 2 : (length(x)-1).

[Jason]

This didnt change anything unless I misunderstood what you were saying. Matlab is beating me down today...

clear;
clc;

load('noisydata.mat','x');
N = x;
y = zeros(1,length(N-1));
y(1) = x(1);
for k = 2:1:(length(x)-1);
    y(k) = (x(k-1)+x(k+1))/2;
    y(length(x))= x(end);
end 

   figure(1)
  plot(N,y)
  xlabel('noisydata')
  ylabel('samp_num_ind')

figure(2)
plot(y(k))
xlabel('smoothedData')
ylabel('sampleNum')

[Image Analyst]

Well I said to put y(length(x))= x(end); after the loop, not inside it.

And for some reason you plotted only a single point

plot(y(k))

instead of the whole curve:

plot(x, 'r*-');
hold on;
plot(y, 'b-', 'LineWidth', 2);

But anyway, try that, because other than that it looks like it should work. But you didn't attach noisydata.mat so I can't test it.

[Jason]

sorry about that I thought I did just attach it to this comment. I really appreciate the help to! thank you. The reason I plotted only one point was I because I wasn't sure what to compare the noisy data with, and what to compare the smoothed data with.

Answer 2

[Geoff Hayes]

Jason - I think that you have correctly interpreted what the question is asking (less the plot) but you may want to reconsider the condition for your if statement

 if (k~= 1 & k~= N)

Remember that k is a scalar but N is a vector/array so this comparison is not what you want. If you want to exclude the first and last elements from x then just extend what you have started with the for statement

 for k = 2:1:length(x)-1

So now k ranges from 2 to one less than the length of x which is exactly which values you need to consider (and so no longer have any need for the condition).

[Jason]

This is what I've got and the first part seems to work now but I dont understant what the question asks when it says to plot the original noisy data vs. "The sample number(index)", vice verse plot in a second window smoothed data vs. sample number". am I missing something here?

clear; clc;

load('noisydata.mat','x');
N = x;
y = zeros(1,length(N));

for k = 2:1:length(x)-1
        y(k) = (x(k-1)+x(k+1))/2;
end 
figure(1)
plot(N,y)
xlabel('noisydata')
ylabel('samp_num_ind')

figure(2)
plot(y(k))
xlabel('smoothedData')
ylabel('sampleNum')

[Geoff Hayes]

Jason - you are being asked to plot the original noisy data versus the sample number. The sample numbers are the indices into your noisy matrix x. How would you plot this?

[Jason]

I'm not sure this whole problem has beat me down. I know its probably something simple