bug related to error message "root not bracketed"
Locked
bug related to error message "root not bracketed"
 
|
Hi,
I came across a bug in Quantlib and not sure where to report. Sorry for the intrusion. in file ./ql/math/solver1d.hpp, the following bug causes error like "root not bracketed: f[2.22045e-16,3] -> [-1.534964e-04,-5.486424e-02]" 185 fxMin_ = f(xMin_); 186 //if (fxMin_ == 0.0) // Bugfix JSHI 187 if (fxMin_ <= accuracy) 188 return xMin_; 189 190 fxMax_ = f(xMax_); 191 //if (fxMax_ == 0.0) // Bugfix JSHI 192 if (fxMax_ <= accuracy) 193 return xMax_; I am using version 0.9.9. Thanks and Regards, Jinping Shi The New Busy is not the old busy. Search, chat and e-mail from your inbox. Get started. ------------------------------------------------------------------------------ ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo _______________________________________________ QuantLib-users mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/quantlib-users |
Re: bug related to error message "root not bracketed"
|
|
On Thu, 2010-06-03 at 18:37 +0000, Jinping Shi wrote:
> in file ./ql/math/solver1d.hpp, the following bug causes error like > > "root not bracketed: f[2.22045e-16,3] -> > [-1.534964e-04,-5.486424e-02]" > > > 185 fxMin_ = f(xMin_); > 186 //if (fxMin_ == 0.0) // Bugfix JSHI > 187 if (fxMin_ <= accuracy) > 188 return xMin_; > 189 > 190 fxMax_ = f(xMax_); > 191 //if (fxMax_ == 0.0) // Bugfix JSHI > 192 if (fxMax_ <= accuracy) > 193 return xMax_; Yes, I remember this. The problem is, in this method we don't know if the accuracy was requested on x of f(x); that differs from solver to solver. If f(xMin_) < accuracy, we can return xMin_ if the accuracy was requested on f(x), since f(xMin_) is close enough to 0; but if it was on x, we can't return as we're not sure that xMin_ is close enough to the actual root X (i.e., that xMin_ - X < accuracy.) The only case in which we can return is when f(xMin_) is precisely 0, because we found the root. So, as of now, it's correct for the solver not to return immediately. To fix the problem, we could add some logic later (when the error message is thrown) to ensure that the missing bracketing is not due to numerical error; or we could add some property to the solvers so that we know whether the accuracy is on x or f(x), and return when possible. Would you mind trying to code that? Thanks, Luigi -- Quote me as saying I was misquoted. -- Groucho Marx ------------------------------------------------------------------------------ ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo _______________________________________________ QuantLib-users mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/quantlib-users |
Re: bug related to error message "root not bracketed"
|
|
My further investigation shows that the problem is somewhere else.
I was using <Forward, Linear> in the boostraping and when the input data is noisy (or too many data points), the forward rate can be negative. So the change should be in /ql/termstructures/iterativebootstrap.hpp 157 // bracket 158 Real min = Traits::minValueAfter(i, ts_->data_);// -2.0; //JSHI, allow negative if using forward linear But when one uses <Discount, LogLinear>, it should be strictly positve. > Subject: Re: [Quantlib-users] bug related to error message "root not bracketed" > From: [hidden email] > To: [hidden email] > CC: [hidden email] > Date: Tue, 8 Jun 2010 11:30:29 +0200 > > On Thu, 2010-06-03 at 18:37 +0000, Jinping Shi wrote: > > in file ./ql/math/solver1d.hpp, the following bug causes error like > > > > "root not bracketed: f[2.22045e-16,3] -> > > [-1.534964e-04,-5.486424e-02]" > > > > > > 185 fxMin_ = f(xMin_); > > 186 //if (fxMin_ == 0.0) // Bugfix JSHI > > 187 if (fxMin_ <= accuracy) > > 188 return xMin_; > > 189 > > 190 fxMax_ = f(xMax_); > > 191 //if (fxMax_ == 0.0) // Bugfix JSHI > > 192 if (fxMax_ <= accuracy) > > 193 return xMax_; > > Yes, I remember this. The problem is, in this method we don't know if > the accuracy was requested on x of f(x); that differs from solver to > solver. If f(xMin_) < accuracy, we can return xMin_ if the accuracy was > requested on f(x), since f(xMin_) is close enough to 0; but if it was on > x, we can't return as we're not sure that xMin_ is close enough to the > actual root X (i.e., that xMin_ - X < accuracy.) The only case in which > we can return is when f(xMin_) is precisely 0, because we found the > root. So, as of now, it's correct for the solver not to return > immediately. To fix the problem, we could add some logic later (when > the error message is thrown) to ensure that the missing bracketing is > not due to numerical error; or we could add some property to the solvers > so that we know whether the accuracy is on x or f(x), and return when > possible. Would you mind trying to code that? > > Thanks, > Luigi > > > -- > > Quote me as saying I was misquoted. > -- Groucho Marx > > The New Busy think 9 to 5 is a cute idea. Combine multiple calendars with Hotmail. Get busy. ------------------------------------------------------------------------------ ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo _______________________________________________ QuantLib-users mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/quantlib-users |
Building Apps on iPhone or Android
|
|
Hi,
I am a student
planning to develope an application using quantlib under iPhone or
Android.
Anything to watch
out and any experience to share?
Furthermore, how
portable is the Java version of quantlib under different
platform?
Two models in
mind, Server based with mobile phone as a front end or simply an App in the
phone
Many Thanks
Billy
------------------------------------------------------------------------------ ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo _______________________________________________ QuantLib-users mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/quantlib-users |
Building Apps on iPhone or Android
|
|
Hi,
------------------------------------------------------------------------------ ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo _______________________________________________ QuantLib-users mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/quantlib-users |
Re: Building Apps on iPhone or Android
|
|
Hi,
Billy Ng wrote > Furthermore, how portable is the Java version of quantlib under > different platform? Quantlib under Java uses SWIG, so you would most probably need to get ql itself ported to the target platform natively... Rgds, Anderas ------------------------------------------------------------------------------ ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo _______________________________________________ QuantLib-users mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/quantlib-users |
Re: Building Apps on iPhone or Android
|
|
Hi Anderas, Thank you for your reply >Quantlib under Java uses SWIG, so you would most probably need to get ql >itself ported to the target platform natively... Has anyone done it before? Is it straight forward or need much fine tuning? There is a JQuantlib http://www.jquantlib.org/index.php/Main_Page which is based on quantlib 0.9.7 It is rewritten instead of a direct translation using SWIG Any comment on this too? Billy Ng -----Original Message----- From: Andreas Spengler [mailto:[hidden email]] Sent: Monday, June 14, 2010 9:56 PM To: [hidden email] Subject: Re: [Quantlib-users] Building Apps on iPhone or Android Hi, Billy Ng wrote > Furthermore, how portable is the Java version of quantlib under > different platform? Quantlib under Java uses SWIG, so you would most probably need to get ql itself ported to the target platform natively... Rgds, Anderas ------------------------------------------------------------------------------ ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo _______________________________________________ QuantLib-users mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/quantlib-users No virus found in this incoming message. Checked by AVG - www.avg.com Version: 9.0.829 / Virus Database: 271.1.1/2936 - Release Date: 06/14/10 14:35:00 ------------------------------------------------------------------------------ ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo _______________________________________________ QuantLib-users mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/quantlib-users |
Re: Building Apps on iPhone or Android
|
|
Hi Billy
I do not know how new you are to this forum. But a month ago the following was posted here. I guess you should try to get in touch with Tawanda. Peter ----------------------------------------------------------------- From: Tawanda Gwena [mailto:[hidden email]] Sent: Saturday, May 01, 2010 3:24 PM To: [hidden email] Subject: [Quantlib-users] A small video of QuantLib on the iPhone I've put up a small video showing QuantLib in action on the iPhone. http://tgwena.blogspot.com/ Work continues on exposing the whole quantlib library intuitively. ------------------------------------------------------------------------------ ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo _______________________________________________ QuantLib-users mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/quantlib-users |
Re: Building Apps on iPhone or Android
|
|
Hi Peter,
Many Thanks and his video is very good. I emailed Tawanda already I joined this group just yesterday. I did search in the archive before posting. However, I mistakenly "selected" the user mailing list before search, which actually means "do not search the selected archive" My apology There seems like no porting to Android yet Billy -----Original Message----- From: Peter Toke Heden Ahlgren [mailto:[hidden email]] Sent: Tuesday, June 15, 2010 3:58 PM To: 'Billy Ng'; [hidden email] Subject: SV: [Quantlib-users] Building Apps on iPhone or Android Hi Billy I do not know how new you are to this forum. But a month ago the following was posted here. I guess you should try to get in touch with Tawanda. Peter ----------------------------------------------------------------- From: Tawanda Gwena [mailto:[hidden email]] Sent: Saturday, May 01, 2010 3:24 PM To: [hidden email] Subject: [Quantlib-users] A small video of QuantLib on the iPhone I've put up a small video showing QuantLib in action on the iPhone. http://tgwena.blogspot.com/ Work continues on exposing the whole quantlib library intuitively. No virus found in this incoming message. Checked by AVG - www.avg.com Version: 9.0.829 / Virus Database: 271.1.1/2936 - Release Date: 06/15/10 02:35:00 ------------------------------------------------------------------------------ ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo _______________________________________________ QuantLib-users mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/quantlib-users |
Re: bug related to error message "root not bracketed"
|
|
In reply to this post by Jinping Shi
On Fri, 2010-06-11 at 14:23 +0000, Jinping Shi wrote:
> My further investigation shows that the problem is somewhere else. > > I was using <Forward, Linear> in the boostraping and when the input > data is noisy (or too many data points), the forward rate can be > negative. So the change should be in > > /ql/termstructures/iterativebootstrap.hpp > > 157 // bracket > 158 Real min = Traits::minValueAfter(i, ts_->data_);// -2.0; //JSHI, > allow negative if using forward linear There's no need to do that (especially since the modification should affect only the forwards, while the code above is generic.) If the problem is negative rates, you can enable them. On Windows, open <ql/userconfig.hpp> and uncomment the line # define QL_NEGATIVE_RATES on Linux and any other platform where you run configure, run it as: ./configure --enable-negative-rates In both cases, recompile the library after the modification. Luigi P.S. Maybe we should enable negative rates by default. People, what do you think? -- The young man knows the rules, but the old man knows the exceptions. -- O. W. Holmes ------------------------------------------------------------------------------ ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo _______________________________________________ QuantLib-users mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/quantlib-users |
Re: bug related to error message "root not bracketed"
|
|
On Thu, Jun 24, 2010 at 11:59 AM, Luigi Ballabio
<[hidden email]> wrote: > P.S. Maybe we should enable negative rates by default. People, what do > you think? Negative rates are an alert that your curve is wrong (besides few extreme cases as JPY very short tenor). A lot of people is stumbling into negative rates just because they have obsolete curve definitions which are not adequate with the current market conditions (multiple Libor curves, collateral discounting, etc) So I'll personally keep negative rates turned off when compiling the library. Anyway we could allow negative rate just to shake off naive users' questions... ;-D ciao -- Nando ------------------------------------------------------------------------------ This SF.net email is sponsored by Sprint What will you do first with EVO, the first 4G phone? Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first _______________________________________________ QuantLib-users mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/quantlib-users |
Quantlib Error: time (30.0037) is past max curve time (30.001)
|
|
In reply to this post by Luigi Ballabio
I am using the sample code swapvaluation.cpp to build termstructure, and made two changes for USD
(1) Calendar swap_calendar = UnitedStates(UnitedStates::GovernmentBond); (2) boost::shared_ptr<IborIndex> swap_FloatingLegIndex(new USDLibor(3*Months) ); There seems some inconsistency between these two and I see the following type error message for certain dates: 20090825, 20090826, 20091222, 20100330, 20100523, 20100504, 20100505, 20100823, 20100824, 20100825 It seems a 30 year swap rate, by using some day count, becomes more than 30 years. My swapr ates are quoted as Thirty360(), as in class SwapRateHelper, and the termstructure day count is using ActualActual(ActualActual::ISDA) as in class PiecewiseYieldCurve<Discount,LogLinear>. Where is the problem and how to correct it? If in (1) I use UnitedKingdom(UnitedKingdom::Exchange), the problem goes away. Is that the right thing to do? run date : 2010-8-25 16:0:00 Swap Settlement date: Friday, August 27th, 2010 running 14040H terminate called after throwing an instance of 'QuantLib::Error' what(): ../../../../utility_sys/bin2/include/ql/termstructures/iterativebootstrap.hpp:199: In function `void QuantLib::IterativeBootstrap<Curve>::calculate() const [with Curve = QuantLib::PiecewiseYieldCurve<QuantLib::Discount, QuantLib::LogLinear, QuantLib::IterativeBootstrap>]': 1st iteration: failed at 10th instrument, maturity August 27th, 2040, reference date August 27th, 2010: termstructure.cpp:84: In function `void QuantLib::TermStructure::checkRange(QuantLib::Time, bool) const': time (30.0037) is past max curve time (30.001) Abort Thanks, Jinping Shi ------------------------------------------------------------------------------ Automate Storage Tiering Simply Optimize IT performance and efficiency through flexible, powerful, automated storage tiering capabilities. View this brief to learn how you can reduce costs and improve performance. http://p.sf.net/sfu/dell-sfdev2dev _______________________________________________ QuantLib-users mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/quantlib-users |
Re: Quantlib Error: time (30.0037) is past max curve time (30.001)
|
|
Found the solution and please ignore the early message. I just need to add one more line:
swapTermStructure->enableExtrapolation(); (Maybe it is worth it to add this line to the example code swapvaluation.cpp) Thanks! From: [hidden email] To: [hidden email] CC: [hidden email] Subject: Quantlib Error: time (30.0037) is past max curve time (30.001) Date: Thu, 9 Sep 2010 14:35:49 +0000 I am using the sample code swapvaluation.cpp to build termstructure, and made two changes for USD (1) Calendar swap_calendar = UnitedStates(UnitedStates::GovernmentBond); (2) boost::shared_ptr<IborIndex> swap_FloatingLegIndex(new USDLibor(3*Months) ); There seems some inconsistency between these two and I see the following type error message for certain dates: 20090825, 20090826, 20091222, 20100330, 20100523, 20100504, 20100505, 20100823, 20100824, 20100825 It seems a 30 year swap rate, by using some day count, becomes more than 30 years. My swapr ates are quoted as Thirty360(), as in class SwapRateHelper, and the termstructure day count is using ActualActual(ActualActual::ISDA) as in class PiecewiseYieldCurve<Discount,LogLinear>. Where is the problem and how to correct it? If in (1) I use UnitedKingdom(UnitedKingdom::Exchange), the problem goes away. Is that the right thing to do? run date : 2010-8-25 16:0:00 Swap Settlement date: Friday, August 27th, 2010 running 14040H terminate called after throwing an instance of 'QuantLib::Error' what(): ../../../../utility_sys/bin2/include/ql/termstructures/iterativebootstrap.hpp:199: In function `void QuantLib::IterativeBootstrap<Curve>::calculate() const [with Curve = QuantLib::PiecewiseYieldCurve<QuantLib::Discount, QuantLib::LogLinear, QuantLib::IterativeBootstrap>]': 1st iteration: failed at 10th instrument, maturity August 27th, 2040, reference date August 27th, 2010: termstructure.cpp:84: In function `void QuantLib::TermStructure::checkRange(QuantLib::Time, bool) const': time (30.0037) is past max curve time (30.001) Abort Thanks, Jinping Shi ------------------------------------------------------------------------------ Automate Storage Tiering Simply Optimize IT performance and efficiency through flexible, powerful, automated storage tiering capabilities. View this brief to learn how you can reduce costs and improve performance. http://p.sf.net/sfu/dell-sfdev2dev _______________________________________________ QuantLib-users mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/quantlib-users |
Re: Quantlib Error: time (30.0037) is past max curve time (30.001)
|
|
On Thu, 2010-09-09 at 15:25 +0000, Jinping Shi wrote:
> Found the solution and please ignore the early message. I just need > to add one more line: > > swapTermStructure->enableExtrapolation(); Yes, that works, but it shouldn't be necessary. It's probably a calendar issue. You're using UnitedStates, while USDLibor uses a joint calendar of UnitedStates and UnitedKingdom. Luigi -- There is no such thing as public opinion. There is only published opinion. -- Winston Churchill ------------------------------------------------------------------------------ Automate Storage Tiering Simply Optimize IT performance and efficiency through flexible, powerful, automated storage tiering capabilities. View this brief to learn how you can reduce costs and improve performance. http://p.sf.net/sfu/dell-sfdev2dev _______________________________________________ QuantLib-users mailing list [hidden email] https://lists.sourceforge.net/lists/listinfo/quantlib-users |
«
Return to quantlib-users
|
1 view|%1 views
| Free forum by Nabble | Edit this page |