copysign

odl.copysign(x1, x2, out=None)

Composes a floating-point value with the magnitude of x1_i and the sign of x2_i for each element of the input array x1.