How can I store a value from each iteration of a for loop, ready for plotting?

Hey guys, I'm trying a fairly basic exercise, to simulate a bouncing ball by calculating it's height. Currently what i have is this:

height = 150
velocity = 0
acceleration = -9.81
bounce_factor = -0.4
increment = 0.01
end_time = 20

hold on

for time = 0 : increment : end_time
    height = height + velocity * increment
    velocity = velocity + acceleration * increment
    if (height <= 0 && velocity < 0)
        height = 0
        velocity = velocity * bounce_factor
    end
    scatter(time,height)
end

This works but ideally I'd like to display using a smooth linegraph. My main approach would be to store the height in an array of equal size to the time variable, and then plot the two. I am having troubles iterating the array, I am more used to C and the old fashioned approach to arrays, and am unsure of how i can store all my values.

I have tried this approac to no avail:

height_store = zeros(end_time/increment,1)
store_inc = 0

hold on

for time = 0 : increment : end_time
    height = height + velocity * increment
    velocity = velocity + acceleration * increment
    if (height <= 0 && velocity < 0)
        height = 0
        velocity = velocity * bounce_factor
    end
    height_store(store_inc) = height
    store_inc = store_inc + 1

end

plot(time,height_store)

This seems really basic, so it's proabably just a matter of properly using Matlabs data types. I checked the documentation and couldn't find what I needed (or didn't recognise it as such!)

Many thanks

Answer 1

[Ganesh Hegade]

you can store values like this.

   height = 150;
    velocity = 0;
    acceleration = -9.81;
    bounce_factor = -0.4;
    increment = 0.01;
    end_time = 20;
    Height = []; 
    for nTimeLen= 1: length(0 : increment : end_time) 
        height = height + velocity * increment ; 
        velocity = velocity + acceleration * increment ; 
        if height <= 0 && velocity < 0
            height = 0; 
            velocity = velocity * bounce_factor; 
        end

        Height(nTimeLen) = height; 
    end
  mTime = 0 : increment : end_time; 
  plot(mTime , Height);

[A]

Thanks for your answer, that was really helpful!

for nTimeLen= 1: length(0 : increment : end_time)

I dont understand what's happening here, can you please explain the syntax of this line?

Thanks again

[Ganesh Hegade]

Here i am running loop from 1 to length of the matrix [0 : increment : end_time] instead of taking each element of the time signal. So just i am getting increment numbers for for loop instead of actual values. Example: now nTimeLen = 1, 2, 3...... till end earlier loop was 0, 0.01, 0.02 .....

[Stephen23]

Note this code has several "features" that could be improved:

• expanding arrays inside loop without array preallocation: for large arrays this is very inefficient as the array must get moved in memory each time it changes size.
• creating a large vector twice: first length(0:increment:end_time), then later mTime = 0:increment:end_time;. Simply create the array once and measure its length.
• Variables whose names only differ by case: although this is permitted it is a regular cause of bugs by beginners (as evidenced by many threads on this forum), and is not recommended.

Answer 2

[Stephen23]

You should preallocate the output array and use indexing to put the values into it. In this example I also create an array for the velocity: this is not strictly required and is a tad slower, but it allows you to generate all time, height, and velocity vectors (e.g. for plotting). I preallocated the arrays using |NaN|s: this is a simple way to check if data is being calculated, and where bugs occur during a loop. Note that the initial values do not change when the code is run: this is very useful when debugging code!

timeV = 0:increment:end_time;
veloV = nan(size(timeV));
highV = nan(size(timeV));
veloV(1) = velocity;
highV(1) = height;
for k = 2:numel(timeV)
    highV(k) = highV(k-1) + increment*veloV(k-1);
    veloV(k) = veloV(k-1) + increment*acceleration;
    if highV(k)<=0 && veloV(k)<0
        highV(k) = 0;
        veloV(k) = veloV(k)*bounce_factor;
    end
end

you can then do a simple plot:

plot(timeV,highV)

or plot the velocity as well:

plotyy(timeV,highV,timeV,veloV)