WeightingPotential: replace hard-coded reference position calculation with model method

 * strictly zero by definition.

 */

double Detector::getWeightingPotential(const ROOT::Math::XYZPoint& local_pos, const Pixel::Index& reference) const {

    auto size = model_->getPixelSize();

    // WARNING This relies on the origin of the local coordinate system

    auto local_x = size.x() * reference.x();

    auto local_y = size.y() * reference.y();

    auto ref = static_cast<ROOT::Math::XYPoint>(model_->getPixelCenter(reference.x(), reference.y()));

    // Requiring to extrapolate the field along z because equilibrium means no change in weighting potential,

    // Without this, we would get large jumps close to the electrode once charge carriers cross the boundary.

    return weighting_potential_.getRelativeTo(local_pos, {local_x, local_y}, true);

    return weighting_potential_.getRelativeTo(local_pos, ref, true);

}

/**