Multi-point soma gets interpreted differently

[michaeldeistler]

1 1 0 0 0 10 -1
2 1 -10 0 0 10 1
3 1 10 0 0 10 1
4 3 -10 -10 0 2 1
5 3 -100 -100 0 2 4

NEURON creates just a single soma section, whereas Jaxley creates two soma branches. I actually find Jaxley more intuitive here...in particular, the NEURON reconstruction changes vastly if one changes the ordering of nodes. E.g.:

1 1 0 0 0 10 -1
2 3 -10 -10 0 2 1
3 3 -100 -100 0 2 3
4 1 -10 0 0 10 1
5 1 10 0 0 10 1

gets put into four sections in NEURON. Jaxley consistently builds the same morphology out of these SWC files.