Investigate using the 'survival' package for signal analysis in quantstrat

[jaymon0703]

Liberally edited from a prior email from @braverock:

"Survival models measure the probability that some event transition will happen in the future.

For signal analysis, the transitions would likely be trading relevant things like pullbacks (so you could us resting limits), up x%, down x%, etc. within some specific time horizon.

This is called a 'competing risks' model, appropriately enough.

Survival models also have a set of 'progression' states (they were developed for modeling disease progression) that could tie up-1%, up-2% etc states together.

Maybe we can get it into quantstrat eventually."