magnitude_of()

template<typename T >

template<typename U >

static double MayaFlux::Kinesis::HampelFilter< T >::magnitude_of ( const U &  v )

inlinestaticprivatenoexcept

Scalar reduction used for the MAD comparison.

Plain arithmetic T reduces to its absolute value directly. glm vector T reduces to its length. This means HampelFilter screens on magnitude of change, not direction: a candidate whose magnitude is consistent with recent history is accepted even if its direction differs, which is deliberate, since a real sharp turn is a direction change with unremarkable magnitude and should not be screened out as if it were a spike.

Definition at line 194 of file HampelFilter.hpp.

    {
        if constexpr (requires { glm::length(v); }) {
            return static_cast<double>(glm::length(v));
        } else {
            return static_cast<double>(std::abs(v));
        }
    }

Referenced by MayaFlux::Kinesis::HampelFilter< T >::accept().

Here is the caller graph for this function: