a little help here guys. im having trouble understanding the method by which the bias is computed. i see that the estimated angle is found using:

kalman->x_angle += dt * (dotAngle - kalman->x_bias)

from my understanding, this would imply that the x_bias is in the units of (angle/second). however, from the update equation i see that the bias is computed via

kalman->x_bias += K_1 * y;

where y is

const float y = angle_m - kalman->x_angle;

. forgive my weak skills in math but i seem to see that the bias is simply a constant (k_1) multiplied by a difference in angle (angle_m - kalman->x_angle). doesnt this mean the bias is already in the units of (angle) which will lead to

kalman->x_angle += dt * (dotAngle) - (kalman->x_bias)

. thank you for your inputs