[MATH-847] Using Rotation to convert Euler angles to Quaternions produces wrong results

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Not A Problem
• Affects Version/s: 3.0
• Fix Version/s: 3.1

Environment

1.6

• Assignee: Unassigned
• Reporter: M@ Dunlap
• Created: 16/Aug/12 18:57
• Updated: 04/Mar/13 18:53
• Resolved: 17/Aug/12 06:17

Description

import org.apache.commons.math3.geometry.euclidean.threed.Rotation;
import org.apache.commons.math3.geometry.euclidean.threed.RotationOrder;

public class temp {
public static void main(String args[])

{ Rotation r = new Rotation(RotationOrder.XYZ, -Math.PI / 2d, 0, 0); System.out.println("(" + r.getQ0() + " " + r.getQ1() + " " + r.getQ2() + " " + r.getQ3() + ")"); }

}

Prints (0.707 0.707 0.0 0.0) (in-sig-figs elided), but when I type the same thing into Wolfram Alpha I get (.707 -.707 0 0) (note the negative) see: http://www.wolframalpha.com/input/?i=euler+angles&a=*C.euler+angles-_*Formula.dflt-&a=*FP.EulerRotation.EAS-_e123&f3=-pi%2F2+rad&f=EulerRotation.th1_-pi%2F2+rad&f4=0&f=EulerRotation.th2_0&f5=0&f=EulerRotation.th3_0

One of the guys in the lab suggested that if Rotation is assuming the Euler angle is in a left-handed coordinate space this is an expected result, but if that's the case the question is, why is the less popular coordinate system the only option?