How to plot a curved signal ?

Can anybody tell me how can I plot a signal like that one represented in the figure below ? I just need the part of the signal from 10 to 15. I have the following code until now :

time = 0:0.1:15;

xArray = zeros(1,numel(time));  %numel gives the number of the elements that an array has

for index = 1:numel(time)
    if(time(index)>5 && time(index)<7)
        xArray(index) = (time(index)-5)/2;
    elseif(time(index))>= 7 && (time(index))<10
        xArray(index) = 1;
    elseif ((time(index) >= 10 && time(index)<15))
        xArray(index) = (15 - time(index))/5;
    end 
end

figure (2)
plot(time,xArray);
axis([0 15 0 5]);

Answer 1

[Image Analyst]

Try this:

numPoints = 1000;
times = linspace(0, 15, numPoints);
xArray = zeros(1,numel(times));  % numel gives the number of the elements that an array has
for index = 1 : numel(times)
    if times(index) > 5 && times(index) < 7
        xArray(index) = (times(index)-5)/2;
  elseif times(index) >= 7 && times(index) < 10
        xArray(index) = 1;
    elseif ((times(index) >= 10 && times(index)<15))
    a = 0.7; % Smaller for flatter, larger for steeper
        xArray(index) = exp(-a * (times(index) - 10));
    end 
end
plot(times, xArray, 'b-', 'LineWidth', 2);
xlim([0, 15]);
ylim([0, 2]);
ax = gca;
ax.YTick = [0 1];
ax.XTick = [5 7 10 15];
ax.FontSize = 15;
grid on;
xlabel('f[Hz]', 'FontSize', 15);
ylabel('x', 'FontSize', 15);
title('xArray vs. Time', 'FontSize', 15);

[Biro Darius]

Thank you very much. I'm new in MATLAB and I'm still learning things. Thanks.

Answer 2

[Akira Agata]

How about the following?

figure
plot(time,xArray)
ax = gca;
ax.YTick = [0 1];
ax.XTick = [5 7 10 15];

Answer 3

[Biro Darius]

I've came back with another question related to this subject. Now I have to build a signal such that it's spectrum looks like that one represented in the picture attached above. I mean I must have a signal and if I want to obtain the spectrum of my signal, his spectrum should looks like that one in the picture. I hope I was pretty clear. Thanks.

[Image Analyst]

Call ifft() on the signal.

[Biro Darius]

Hi, I've tried before but it's not what I needed. My teacher suggested that I should build a sum of sinusoids for each part of the figure ( a sum of sinusoids from 5 to 7, another sum from 7 to 10 and the last sum from 10 to 15). He also suggested to use some iterations for frequency and amplitude. After all of this I should be able to build a unique signal consisting of the sum of the three sums of sinusoides and after I will call fft() on the final signal I should obtain a spectrum which has almost the same pattern like that one from the image. I hope I was pretty clear. Thanks.

[Image Analyst]

Yes, he wants you to do it "manually" instead of having the ifft() function do it for you.