Tracking with wrong sign for R_ref gives NaN with no message

[DavidSagan]

Test program:

using BeamTracking
using Beamlines

@eles begin
  q = Drift(L = 1, tracking_method = Exact())
end

bb = Beamline([q], R_ref = -1e7, species_ref = Species("electron"))

v1 = [0.0  0.0  0.0  0.0  0.0  0.0]

b1 = Bunch(deepcopy(v1), species=Species("electron"), R_ref = -1e7)
track!(b1, q)
println(b1.coords.v)

b2 = Bunch(deepcopy(v1), species=Species("electron"), R_ref = 1e7)
track!(b2, q)
println(b2.coords.v)

Output is:

julia> include("local/z.jl")
[0.0 0.0 0.0 0.0 0.0 0.0]
[0.0 0.0 0.0 0.0 NaN 0.0]

[DavidSagan]

Also:

using BeamTracking
using Beamlines

@eles begin
  q = Drift(L = 1, tracking_method = Exact())
end

bb = Beamline([q], R_ref = -1e7, species_ref = Species("electron"))

v1 = [0.0  0.0  0.0  0.0  0.0  0.0]

b1 = Bunch(deepcopy(v1), species=Species("electron"))
track!(b1, q)
println(b1.coords.v)

Gives:

julia> include("local/z.jl")
[0.0 0.0 0.0 0.0 NaN 0.0]

[mattsignorelli]

OK for now, instead of track!(b1, q), do track!(b1, bb). Basically for the Beamline level track! there is this automatic check done to set reference energies + species equal, but it hasn't been put in the the LineElement level track!, which is what you call here. While it would be easy to use the same code for the element-level track, we will have to change it anyways to properly handle ramping in a more generic framework, so I don't want to add code to later have to remove it

[DavidSagan]

@mattsignorelli There still is the problem that with R_ref with the wrong sign in a Bunch the tracking is off.