Not the kind of news I like to disseminate late Friday afternoon, but there
seems to be a problem (sometimes) with the helical extrapolation, I think
some near relation to the problem Ray had with his fortran version. In
particular, I suspect I'm (sometimes) getting a "shortest path length
to escape volume" corresponding to going backwards in time when I'm trying
to find path length to reach a particular radius. The length I actually
use to propagate the particle is the absolute value of this, but this
isn't right either; I won't end up at the boundary I'm expecting to
end up at.
There were a couple other small problems I found and fixed, but
it didn't seem worthwhile making another distribution until the thing
really works.
Joanne
|