Sum of probabilities is > 1 for some short sequences

[Max Ward]

Sum of probabilities for all structures of UAGCUAGCUUGAUC = 1.001911832338799.

This was computed by the following.

(1) Enumerating all possible secondary structures.
(2) Using efn to compute their energy. (Note that this should underestimate since it only takes the optimal CTD arrangement)
(3) Calling Boltzmann on the energies, then divide by Q to get a probability.
(4) Sum probabilities.

I don't have a (simple) code example yet since I compute all this via SWIG constructed python bindings inside a script.

[Edgeworth]

Can you check if this still happens after 0e478fd ?

[Max Ward]

Seems to be fixed now.