remove_jumps

ahrs.common.quaternion.QuaternionArray.remove_jumps(self) → None

Flip sign of negated quaternions.

Some estimations and measurements of quaternions might have “jumps” produced when their values are multiplied by -1. They still represent the same rotation, but the continuity of the signal “flips”, making it difficult to evaluate continuously.

To revert this, the “flipped” instances are identified and multiplied by -1 to “flip them back”. This function does that correction over all values of the attribute array.

Examples

>>> qts = np.tile([1., -2., 3., -4], (5, 1))    # Five equal arrays
>>> v = np.random.randn(5, 4)*0.1               # Gaussian noise
>>> Q = QuaternionArray(qts + v)
>>> Q.view()
QuaternionArray([[ 0.17614144, -0.39173347,  0.56303067, -0.70605634],
                 [ 0.17607515, -0.3839024 ,  0.52673809, -0.73767437],
                 [ 0.16823806, -0.35898889,  0.53664261, -0.74487424],
                 [ 0.17094453, -0.3723117 ,  0.54109885, -0.73442086],
                 [ 0.1862619 , -0.38421818,  0.5260265 , -0.73551276]])
>>> Q[1:3] *= -1    # 2nd and 3rd Quaternions "flip"
>>> Q.view()
QuaternionArray([[ 0.17614144, -0.39173347,  0.56303067, -0.70605634],
                 [-0.17607515,  0.3839024 , -0.52673809,  0.73767437],
                 [-0.16823806,  0.35898889, -0.53664261,  0.74487424],
                 [ 0.17094453, -0.3723117 ,  0.54109885, -0.73442086],
                 [ 0.1862619 , -0.38421818,  0.5260265 , -0.73551276]])
>>> Q.remove_jumps()
>>> Q.view()
QuaternionArray([[ 0.17614144, -0.39173347,  0.56303067, -0.70605634],
                 [ 0.17607515, -0.3839024 ,  0.52673809, -0.73767437],
                 [ 0.16823806, -0.35898889,  0.53664261, -0.74487424],
                 [ 0.17094453, -0.3723117 ,  0.54109885, -0.73442086],
                 [ 0.1862619 , -0.38421818,  0.5260265 , -0.73551276]])