Trimming a curve - how to find nearest number and replace

I have the x and y coordinates for a bathymetric curve eg...

x   y
0   1.36
0.03  -0.92
0.13  -1.68
1.69  -3.72
6.63  -3.35
9.18  -3.36
20.03  -3.26
31.75  -2.93
40.15  -1.70
46.21   0.10
69.97   2.26

I would like to trim this curve to a certain level line with at y=-0.15 and x from 0 to 45 I need to both replace the first coordinate with (0,-0.15) and make the final coordinate (45,-0.15) removing any x,y values greater than this.

I have this for a large number of data sets so of course a loop would be preferential but I cannot work out how to do these operations through code alone. The image below might clarify what I mean a little more, though it was made with different numbers.

Thanks in advance for any help

Answer 1

[Greg Dionne]

Try:

myLimit = -0.15;

iFirst = find(y<myLimit, 1, 'first');
iLast = find(y<myLimit, 1, 'last');

xdesired = x(iFirst:iLast);
ydesired = y(iFirst:iLast);

plot(x, y); 
hold on
plot(xdesired, ydesired);
legend('Original','Trimmed')

[Dom Smith]

That's certainly close, it clears the values I don't want.. How then could I insert the new coordinates (0,mylimit) at the start and (45,mylimit) at the end?

[Dom Smith]

With a bit of modification I've made this work thank you

[Greg Dionne]

OK. good.

For the concatenation I would probably do something like:

% use "," instead of ";" if xdesired and ydesired are row vectors
xdesired = [0; xdesired; 45];
ydesired = [mylimit; xdesired; mylimit];

Glad you got past this.

-G

Answer 2

[Image Analyst]

The data you supplied does not look like your plots. But to clamp data to some level, try this:

xy = [...
0   1.36
0.03  -0.92
0.13  -1.68
1.69  -3.72
6.63  -3.35
9.18  -3.36
20.03  -3.26
31.75  -2.93
40.15  -1.70
46.21   0.10
69.97   2.26];
x = xy(:, 1);
y = xy(:, 2);
subplot(2, 1, 1);
plot(x, y, 'b-');
grid on;
yClamped = y; % Initialize
clampLevel = 0.15;
yClamped(y>clampLevel) = clampLevel; % Do the clamping
subplot(2, 1, 2);
plot(x, yClamped, 'b-');
grid on;

[Dom Smith]

With a bit of modification I made the answer above work for my needs, but thanks so much for your help!