NAN error

Unity Engine
1

Hi, anyone know how I can fix this error?

rigidbody.velocity assign attempt for ‘Item(Clone)’ is not valid. Input velocity is { NaN, NaN, NaN }.

It works most of the time but sometimes I get this error. I’m normalizing the direction before I multiply it with the force.

Basically I’m setting the velocity to the output of this function:

    private Vector3 SimulateProjectile(Vector3 target)
    {
        Rigidbody.drag = 0f;
        Rigidbody.useGravity = true;

        float firingAngle = 45.0f;

        var dir = target - startTransform.position;  // get target direction
        var h = dir.y;  // get height difference
        dir.y = 0;  // retain only the horizontal direction
        var dist = dir.magnitude;  // get horizontal distance
        var a = firingAngle * Mathf.Deg2Rad;  // convert angle to radians
        dir.y = dist * Mathf.Tan(a);  // set dir to the elevation angle
        dist += h / Mathf.Tan(a);  // correct for small height differences
                                   // calculate the velocity magnitude
        var vel = Mathf.Sqrt(dist * Physics.gravity.magnitude / Mathf.Sin(2 * a));
        return vel * dir.normalized;
    }
1 Like
2

Maybe you’re producing a NaN due to you dividing by zero in lines 14 or 16 above?

1 Like
3

I printed out the values of those equations to make sure and they’re never 0. I get the error when both of those values are 1. I mean the firing angle is always the same - I’m setting it to 45 in that code so those tan and sin 2 * a can’t be 0.

4

It should be a simple matter to debug the script by adding the following then tracing back what is actually infinity/nan.

        if (float.IsNaN(value))
        {
            Debug.Break();
        }

Check that the output of the function is indeed returning this. Set a breakpoint then it should be easy to inspect which part is infinity.

1 Like
5

Using that I found out that at this point, it becomes NaN:
var vel = Mathf.Sqrt(dist * Physics.gravity.magnitude / Mathf.Sin(2 * a));

That’s because dist * Physics.gravity.magnitude / Mathf.Sin(2 * a) is negative.

I need to find some new calculation that works on large height differences (now I understand that comment). Unfortunately the ones I’ve found don’t work for some reason.

6

Well, the square root of a negative number is an imaginary number which will result in NAN. Use Mathf.Abs and (if directionaility is important``) ``Mathf.Sign instead:

Mathf.Sqrt(Mathf.Abs(dist) * Physics.gravity.magnitude / Mathf.Sin(2f * a)) * Mathf.Sign(dist);

2 Likes
7

^ Yes, exactly as above! Thanks @grizzly

1 Like
8

Is that correct? I tried it and the object seems to be moving in the opposite direction when dist is negative.

Doing * -Mathf.Sign(dist) doesn’t seem to be correct either. Sorry, I’m not great at math but I’m trying to understand.

9

If directionality (whether distance is above or below the target) isn’t important then you only need the first part;

Mathf.Sqrt(Mathf.Abs(dist) * Physics.gravity.magnitude / Mathf.Sin(2f * a));

1 Like
10

But directionality is important. Getting the absolute value was the first thing I tried after I realized that dist being negative was causing these problems and the result was worse than NAN.

11

If you’re experiencing a separate issue please provide more information. What’s worse than NAN and what exactly, result were you expecting?