Re: [Quantlib-dev] Can anyone explain/provide the numerical algorithm used in SurvivalProbability trait?

[Luigi B. <lui...@gm...>]

As far as I can make out of it:
- for the guess: if we have a previous curve (i.e., if validData is true),
the guess is its value for the node we're bootstrapping; if not, and if
we're at the first node (i == 1), it estimates the survival probability
from a hazard rate of 1% over a three-months period (0.25), which is just a
guess from some sensible initial values; if we're at a later node, then we
have already built part of the curve and we get a guess by extrapolating it.
- for the min value, I'm not sure of the idea behind the implementation.
If we have a previous curve, it seems to assume that the survival
probability can't be less than half the previous value; if not, it starts
from the last node and calculates the minimum survival probability based on
some maximum hazard rate.  It might be simpler to just take a minimum value
a bit above zero; but then again, this is just to limit the search range of
the solver and probably doesn't make a difference for the final value.

Luigi


On Mon, Jan 29, 2018 at 12:57 AM yuzhao88 <yuz...@gm...> wrote:

> Anyone has any idea about the numerical methods used?
>
> On Mon, Jan 22, 2018, 10:04 PM yuzhao88 <yuz...@gm...> wrote:
>
>> I am trying to make sense of the guess method in the SurvivalProbability
>> trait. Can anyone explain/provide some reference of the parameters and the
>> algorithms used in the guess and minValueAfter function?
>>
>> Thanks in advance!
>>
>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> _______________________________________________
> QuantLib-dev mailing list
> Qua...@li...
> https://lists.sourceforge.net/lists/listinfo/quantlib-dev
>