Throw 5 Gev muons into the N arm:
cfmgmc.input = (5. 5.0 18.0 90.0 14.0 100 0.0 0.0 0.0 0.0)


From rcas cvs3 files fvtx_mc_eval_20perc4.root, fvtx_mc_eval_50100_4.root, fvtx_mc_eval_perf4.root.

x0reco	track projection to the z=0 plane
y0reco	track projection to the z=0 plane
z0reco	z=0 or to z-axis

5.0 GeV μ	x0reco		y0reco		z0reco
	mean	sigma	mean	sigma	mean	sigma
pi = poriginal	~0	47	~0	54	-	-
pi->pi±20%	~0	131	~0	71	-	-
px=py=50, pz=100	0.08	228	~0	354	-	-
straight-line fit	~0	231	~0	356	~0	166

Note that with perfect information, the Kalman filter projects to %50μm, which is roughly what we expect from multiple scattering.

On the other extreme, when no momentum information is available, and the best we can do is assume a straight line by feeding in very high momenta, the Kalman filter returns the same values as a straight-line fit.

The realistic case is one where we have partial knowledge of the momentum, simulated by smearing the input momentum by 20%.