Re: [Quantlib-users] Refactoring of voltermstructure.hpp

[Ferdinando A. <na...@qu...>]

>I'm trying to refactor the file voltermstructure.hpp
as the author of that file I'm pleased to welcome you. Please just allow me 
half a day to check if I have major modifications/additions to that file 
that I haven't checked into the CVS yet.

>Also, I would like to move part of the code in a new file
>voltermstructure.cpp
no problem at all.

>I do not understand the interface of the class
>LocalVolTermStructure, e.g.,
>       public:
>         double localVol(const Date& date1,
>                         const Date& date2,
>                         double underlyingLevel,
>                         bool extrapolate = false) const;
>and similarly for the others.
>
>Can I drop the extra parameter?
You can. I cannot figure out what I had in mind when I added that second 
date. Probably it was just a wrong cut&paste.

>I'd like to add few methods to compute the derivatives
>of BlackVolTermStructure. These are needed to implement the method
>described in Rebonato(Vol and Cor.) at chapter 6.
>The added methods are
>
>virtual double timeDerivative(...)
>virtual double strikeDerivative(...)
>virtual double strikeSecondDerivative(...)
>
>and they would be implemented as finite difference in the
>BlackVolTermStructure class.
that would be great, and it was in my original plan, if only I had time 
enough to implement it.

While on this subject: do you plan to contribute a constructor for a 
LocalVolTermStructure class that uses a BlackVolTermStructure as input to 
bootstrap the local vol surface?
In that case I would consider worthwhile to pay attention to degenerate 
cases as 1) constant Black vol (no time/strike dependencies) 2) time 
dependent Black vol (no strike dependencies): while I do understand that 
these cases do not have real-life values, I think they help tremendously 
when it comes to testing and didactical issues.
Of course in these 2 cases there is no real difference between a local vol 
curve and the corresponding Black vol curve, it's just a matter of 
providing a local vol interface to what is basically a simple Black vol curve.
Am I missing something? Do you agree with my comments?

thank you for your help

ciao -- Nando





ciao -- Nando


>Marco Marchioro.
>
>
>
>
>
>-------------------------------------------------------
>This SF.net email is sponsored by: Get the new Palm Tungsten T handheld. 
>Power & Color in a compact size! 
>http://ads.sourceforge.net/cgi-bin/redirect.pl?palm0002en
>_______________________________________________
>Quantlib-users mailing list
>Qua...@li...
>https://lists.sourceforge.net/lists/listinfo/quantlib-users