fzero help

y=@(T) I*T*sind(A)-(0.5*G*(T^2));
root = fzero(y, 0)

abouve eqation is for a projectile that projected at an angle (gose in a kind of ar half oval shape)

want to find the time it hist ground but fzero only gives me the biginning which is 0

[Walter Roberson]

Maybe it just doesn't like buxZED... won't let me submit an answer to this one.

[Walter Roberson]

root = fzero(y,[eps, realmax])

[buxZED]

any clue why?

can you post as a comment?

[Paulo Silva]

Hey Walter, answers stop working for me too :(

We're sorry, but something went wrong.

We've been notified about this issue and we'll take a look at it shortly.

[buxZED]

??? Error using ==> fzero at 260

Function values at interval endpoints must be finite and real.

Error in ==> p34 at 26

root = fzero(y,[eps, realmax])

>>

[buxZED]

root = fzero(y,eps, realmax)

gives me a - answer, imposible

[Matt Fig]

Plot the function, look at the plot and pick a closer guess than zero.

[buxZED]

@matt Fig

that was my inicial thought

but we are not alowed to use neumerical methods or guess work

[Matt Fig]

You do realize that FZERO is a numerical root finder, right?

[buxZED]

is there a way to force fzero to go to the next positive root?

[Walter Roberson]

Ummm, make it fzero(y,[eps,sqrt(realmax)/G]

Using fzero(y,eps,realmax) would cause realmax to be interpreted as an Option, and then it would use eps as the starting guess, not constrained from going backwards.

[Walter Roberson]

No, there is no way to force fzero to go to the next positive root. You can only exclude the area you already searched from the area fzero is to search next.

[Matt Fig]

rt = 0;

d = .1;

cnt = 1;

while ~rt

rt = fzero(y,rt+d*cnt);

cnt = cnt + 1;

end

[buxZED]

it works

but i want to understand what while ~rt means (what dose ~ say)

dose this stop counting as it finds the first root after zero?

[Walter Roberson]

~ is logical negation. It is true if and only if the thing being tested is exactly 0.

[buxZED]

so dose this stops finding roots after it reached the 1st root after 0?

[Matt Fig]

When rt is not zero, then ~rt is false. That would stop the WHILE loop.

[buxZED]

y=@(T) I*T*sind(A)-(0.5*G*(T^2));

%root = fzero(y, eps, realmax)

rt = 0;

d = .1;

cnt = 1;

while ~rt

rt = fzero(y,rt+d*cnt);

cnt = cnt + 1;

end

fprintf('the root is %g\n' , rt)

gives me the root is 7.11215e-026

but its not posible as the projectile takes around 1.5sec to reach maximum hight, the root should be around 3

can you point me in the right derection?

[buxZED]

is there a way to systematicaly increase the T value till Y become zero and get that value?

[Matt Fig]

Change from:

while ~rt

to:

while rt<.1

or similar. It might help if you actually gave us some values for I, A and G.

[buxZED]

it worked Matt

thank you so much

and also ~ would allow me to use matlab in many new ways

thank you again :D

[Matt Tearle]

Can I suggest using an initial guess of 2*I*sind(A)/G? Just sayin'...

[Matt Fig]

Come on now, Matt! Where is the fun of solving the equation with your brain instead of making MATLAB do it?

Answer 1

[Honglei Chen]

The anwser is in the comments above

From Matt Fig

rt = 0; 
d = .1; 
cnt = 1; 
while rt<.1 
rt = fzero(y,rt+d*cnt); 
cnt = cnt + 1; 
end

[Matt Fig]

Digging deep, Honglei! How did you come across this one from over a year ago?

[Honglei Chen]

I'm helping cleaning up some old, unanswered questions :)