ffts-develop Mailing List for FFTS
Status: Planning
Brought to you by:
mghiggins
You can subscribe to this list here.
2000 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
(1) |
Sep
(2) |
Oct
(1) |
Nov
|
Dec
|
---|---|---|---|---|---|---|---|---|---|---|---|---|
2001 |
Jan
|
Feb
|
Mar
(3) |
Apr
(1) |
May
(6) |
Jun
(1) |
Jul
|
Aug
(1) |
Sep
(4) |
Oct
(3) |
Nov
(4) |
Dec
(1) |
From: Ferdinando A. <fer...@am...> - 2001-12-06 13:49:30
|
Hi all what happened to FFTS? No releases available from SourceForge (http://sourceforge.net/projects/ffts/), nothing in FreshMeat ... ciao -- Nando |
From: Mark H. <mgh...@ya...> - 2001-11-10 18:20:23
|
I've just released v0.83 of the Finicky Financial Trading System. The new version eliminates the old Account Manager, which is replaced by a shiny new Portfolio Manager. This tool lets you manage the portfolio hierarchy as well as create/manage accounts (accounts are where trades are booked; they live in portfolios). I also added the ability to limit the types of underlying asset you can trade in an account. For example, I kept accidentally booking MSFT options in my GS account; now if I try that, the trade entry will pop up an error message telling me I can't. I upgraded the vol surface editor a bit. Now you can view bid or ask implied vols as well as midmarket ones. Also, instead of viewing vol by strike, you can view vol by Delta. This is parameterized by ATM vol, skew, and smile, so just three numbers per tenor instead of vols for each traded strike. http://ffts.sourceforge.net ===== ------------------------------------------------------- Mark Higgins, mgh...@ya... http://www.molala.net Opinions expressed herein do not reflect my own views; I haven't had free will since last year when aliens ate my brain. __________________________________________________ Do You Yahoo!? Find a job, post your resume. http://careers.yahoo.com |
From: Mark H. <mgh...@ya...> - 2001-11-06 18:33:07
|
> Note 1: > When I tried to build FFTS, it kept complaining about missing > headers and libraries. Thanks for letting me know! Great info to have. A couple of follow-up questions: 1) What platform are you building it on? 2) What were the missing headers/libraries? MakeVars should contain all the platform-specific info, so maybe I need some kind of configure script that checks the platform and updates MakeVars appropriately. > Note 2: > Looks like Mr. Higgins is not using some IDE Just NEdit (http://www.nedit.org) using a tab width of 3. It's a nice little text editor. > I just imported the root FFTS directory, KDevelop generated > the project file, and I was able to work on the project as if > it was completely built with KDevelop. Very impressive. > Nice work Mr. Higgins Wow! I certainly didn't have KDevelop in mind - but I'll take credit for it anyway. :) I suspect kudos are more due to KDevelop people than me, though. > Note 3: ... historical option data ... Check out http://prdownloads.sourceforge.net/ffts/MarketDB.out.gz This is a dump of the postgresql database I keep my market data in, which I update every now and then. It's got historical option prices for a handful of stocks going back to Jan 00. There are some big gaps, but there's still a lot of juicy info. AFAIK, there's no place that gives out free historical option prices, so I've been forced to create them myself. ===== ------------------------------------------------------- Mark Higgins, mgh...@ya... http://www.molala.net Opinions expressed herein do not reflect my own views; I haven't had free will since last year when aliens ate my brain. __________________________________________________ Do You Yahoo!? Find a job, post your resume. http://careers.yahoo.com |
From: Tiberiu S. <ts...@cs...> - 2001-11-06 18:13:45
|
Hello all, I don't know if this is the right place to post the following. I'll give it a try, though. Note 1: When I tried to build FFTS, it kept complaining about missing headers and libraries. To overcome this I had to make small changes to the MakeVars file: add: OTHER=-L/usr/local/some_package/lib -I/usr/local/some_package/include and then change BASESFX to: BASESFX = $(FFTSDEBUG) $(DBBIT) $(FBIT) $(OTHER) $(TEMPDPTH) $(OPTFLAGS) $(WARNINGS) Note 2: Looks like Mr. Higgins is not using some IDE, but I tried KDevelop, and it worked just fine. I just imported the root FFTS directory, KDevelop generated the project file, and I was able to work on the project as if it was completely built with KDevelop. Very impressive. Nice work Mr. Higgins Note 3: I will try to build a small module to play around with option data, so if Mr higgins or anyone else has historical option data and is willing to share it, please send me a mail. Thanks A Lot T. Stef |
From: Mark H. <mgh...@ya...> - 2001-11-03 02:55:46
|
Right now in the FFTS, all you can do is view the implied vol surface (with the vol surface viewer). It'd be nice to be able to manipulate the vol surface, but we need some small-parameter way of specifying it. I'm thinking of parameterizing the vols for a given tenor by ATM vol, "smile", and "skew", where ATM == forward price, and smile and skew are defined (for a given tenor) as: u = European call Delta - ATM European call Delta Vol( u ) = ATM Vol + skew * u + smile * u^2 + offset( u ) and choosing skew & smile to minimize the sum of offset(u)'s squared for the options that trade in the market. Does anyone reading this have practical experience in the equity derivatives market - how do they specify the vol surface? In strike space or in Delta space? Delta space (as above) seems more theoretically pleasing, but strike space lines up with how the options trade, so it's not clear to me. Perhaps a simpler scheme than the one above is better, since that's fairly numerically intensive. Any ideas? Doing this sort of small-parameter fit will help not only for playing with the market state so see how it affects your risk, but also because it might give some insight into how the markets work and lead to improved modeling. This has certainly been true in the FX vol market, so I suspect building intuition here can go a long way as well. ===== ------------------------------------------------------- Mark Higgins, mgh...@ya... http://www.molala.net Opinions expressed herein do not reflect my own views; I haven't had free will since last year when aliens ate my brain. __________________________________________________ Do You Yahoo!? Find a job, post your resume. http://careers.yahoo.com |
From: Haifa T. <Ha...@he...> - 2001-10-12 15:12:53
|
Hi, As you recommended me,I have insatlled gcc2.95.3 on my sun box. Then I retry to install FFTS. But problems persists. When launching MakeAll, I receive a lot of erros kind of: In file included from /opt/app/usr/local/pgsql/include/libpq++.h:27, from CDatabase.h:30, from CDatabase.cpp:22: /opt/app/usr/local/pgsql/include/pgconnection.h:86: syntax error before'(' .... while the 86th ligne of pgconnection.h looks all right: " string IntToString(int)" Is this due to the postgresql or to problems with string type, I informe you, thas during installation of postgresql I encountered some problems, so I decided to install binary version. What do you recommend me. I'm waiting for your help. Thanks a lot. Haifa. -----Message d'origine----- De : Mark Higgins [mailto:mgh...@ya...] Envoye : mardi, 18. septembre 2001 16:20 A : Haifa Touati Objet : Re: [Ffts-develop] installation problems Thanks for the note! I haven't heard of anyone trying to install it on a sun box yet, so this is new territory. :) To start with, I'd grab the most recent version from the CVS repository... follow the instructions at http://sourceforge.net/cvs/?group_id=9699 to pull it all down. The advantage here is that I've generalised MakeVars a bit, and that might help us out in resolving this. My first guess is that the problem arises because of namespaces... versions of gcc before 3.0 don't require you to explicitly state the namespaces you're using, like std:: which contains string. Try compiling this bit of code and let me know if it gives you the same error: #include <iostream> #include <string> int main( void ) { string s( "hello" ); cout << s << "\n"; return 0; } If my hypothesis is correct, this code will give you the same error message, but adding the line using namespace std; before int main( void ) will make the error go away. If that's the case, it'll require a fair bit of work to solve your problem, since a tonne of source code files will need to be edited. Your best bet for getting the FFTS to work, then, would be to download gcc and install it on your sun box - I'm pretty sure it's available and know it's free. The best version of gcc to get is probably 2.96. --- Haifa Touati <Ha...@he...> wrote: > Hi, > I'm installing FFTS v 0.5 on a Sun 2.8 station. > I have installed Qt, PostgreSQL and LibCurl as mentionned in > the > installation guide. > I have changed the Makefile to use CC sparc compiler instead > of the g++ one > (because this latter isn't very well installed on my station). > when I lunch the MakeAll, I receive errors like : > type name expected instead of string. > So I added in the MakeVars file the stl lib "string.h" > but errors persist. > How can I resolve this? > Does FFTS installation requires g++? > Thanks in advance. > Haifa. > > > _______________________________________________ > ffts-develop mailing list > fft...@li... > https://lists.sourceforge.net/lists/listinfo/ffts-develop ===== ------------------------------------------------------- Mark Higgins, mgh...@ya... http://www.molala.net Opinions expressed herein do not reflect my own views; I haven't had free will since last year when aliens ate my brain. __________________________________________________ Terrorist Attacks on U.S. - How can you help? Donate cash, emergency relief information http://dailynews.yahoo.com/fc/US/Emergency_Information/ |
From: Mark H. <mgh...@ya...> - 2001-10-11 12:49:51
|
I released version 0.82 of the Finicky Financial Trading System today. http://ffts.sourceforge.net The binary rpm is available for download at http://ffts.sourceforge.net/RPMs.html The main changes from 0.81 are: * Proxy support for downloading. Thanks to Dirk Eddelbuettel for figuring out how this works in libCURL and doing testing for me! You can now set proxy settings from the main app, Edit|Preferences menu. * Two new reports: Vega by spot (does a Vega report at different spot levels) and Vega by vol (ditto for different vol levels). Using these you can examine the convexity of the portfolio value to volatility, which gives you some sense of the premium to attach to stochastic volatility. * Better trade handling. You can now "delete" trades in Trade Viewer; also, trades are marked with the user who entered them, for audit purposes. ===== ------------------------------------------------------- Mark Higgins, mgh...@ya... http://www.molala.net Opinions expressed herein do not reflect my own views; I haven't had free will since last year when aliens ate my brain. __________________________________________________ Do You Yahoo!? Make a great connection at Yahoo! Personals. http://personals.yahoo.com |
From: Mark H. <mgh...@ya...> - 2001-10-01 00:35:11
|
I've put out a new release of the FFTS, version 0.81. The changes in this version basically make it easier to install: it now uses more standard makefiles, does a proper make install to /usr/local/FFTS, etc. I've also built source and binary rpm's for Red Hat 7.x. http://ffts.sourceforge.net/ ===== ------------------------------------------------------- Mark Higgins, mgh...@ya... http://www.molala.net Opinions expressed herein do not reflect my own views; I haven't had free will since last year when aliens ate my brain. __________________________________________________ Do You Yahoo!? Listen to your Yahoo! Mail messages from any phone. http://phone.yahoo.com |
From: Mark H. <mgh...@ya...> - 2001-09-23 05:05:27
|
I've released version 0.8 of the Finicky Financial Trading System. You can download the new source tarball from http://ffts.sourceforge.net This version introduces a new application for viewing and editing trades, new risk reports (a dividend rate sensitivity report and a report showing sensitivity to the volatility skew), and several bug fixes. I've also reworked the front end to the html documentation somewhat. __________________________________________________ Do You Yahoo!? Get email alerts & NEW webcam video instant messaging with Yahoo! Messenger. http://im.yahoo.com |
From: Haifa T. <Ha...@he...> - 2001-09-18 17:37:35
|
Thanks a lot for your reply, Exactly, your hypothesis is correct, When I've compiled the test code I received: "Error string not found" and after adding "using namespaces std;" the error goes away. So I will install gcc. Hoping this will solve the problem. Thanks. Haifa. -----Message d'origine----- De : Anis Benbrahim=20 Envoy=E9 : Tuesday, September 18, 2001 5:07 PM =C0 : Haifa Touati Objet : TR: [Ffts-develop] installation problems -----Message d'origine----- De : Mark Higgins [mailto:mgh...@ya...] Envoye : mardi, 18. septembre 2001 16:20 A : Haifa Touati Objet : Re: [Ffts-develop] installation problems Thanks for the note! I haven't heard of anyone trying to install it on a sun box yet, so this is new territory. :) To start with, I'd grab the most recent version from the CVS repository... follow the instructions at http://sourceforge.net/cvs/?group_id=3D9699 to pull it all down. The advantage here is that I've generalised MakeVars a bit, and that might help us out in resolving this. My first guess is that the problem arises because of namespaces... versions of gcc before 3.0 don't require you to explicitly state the namespaces you're using, like std:: which contains string. Try compiling this bit of code and let me know if it gives you the same error: #include <iostream> #include <string> int main( void ) { string s( "hello" ); cout << s << "\n"; return 0; } If my hypothesis is correct, this code will give you the same error message, but adding the line using namespace std; before int main( void ) will make the error go away. If that's the case, it'll require a fair bit of work to solve your problem, since a tonne of source code files will need to be edited. Your best bet for getting the FFTS to work, then, would be to download gcc and install it on your sun box - I'm pretty sure it's available and know it's free. The best version of gcc to get is probably 2.96. --- Haifa Touati <Ha...@he...> wrote: > Hi, > I'm installing FFTS v 0.5 on a Sun 2.8 station. > I have installed Qt, PostgreSQL and LibCurl as mentionned in > the > installation guide. > I have changed the Makefile to use CC sparc compiler instead > of the g++ one > (because this latter isn't very well installed on my station). > when I lunch the MakeAll, I receive errors like : > type name expected instead of string. > So I added in the MakeVars file the stl lib "string.h" > but errors persist. > How can I resolve this? > Does FFTS installation requires g++? > Thanks in advance. > Haifa.=20 >=20 >=20 > _______________________________________________ > ffts-develop mailing list > fft...@li... > https://lists.sourceforge.net/lists/listinfo/ffts-develop =3D=3D=3D=3D=3D ------------------------------------------------------- Mark Higgins, mgh...@ya... http://www.molala.net Opinions expressed herein do not reflect my own views; I haven't had free will since last year when aliens ate my brain. __________________________________________________ Terrorist Attacks on U.S. - How can you help? Donate cash, emergency relief information http://dailynews.yahoo.com/fc/US/Emergency_Information/ |
From: Mark H. <mgh...@ya...> - 2001-09-18 15:51:36
|
Thanks for the note! I haven't heard of anyone trying to install it on a sun box yet, so this is new territory. :) To start with, I'd grab the most recent version from the CVS repository... follow the instructions at http://sourceforge.net/cvs/?group_id=9699 to pull it all down. The advantage here is that I've generalised MakeVars a bit, and that might help us out in resolving this. My first guess is that the problem arises because of namespaces... versions of gcc before 3.0 don't require you to explicitly state the namespaces you're using, like std:: which contains string. Try compiling this bit of code and let me know if it gives you the same error: #include <iostream> #include <string> int main( void ) { string s( "hello" ); cout << s << "\n"; return 0; } If my hypothesis is correct, this code will give you the same error message, but adding the line using namespace std; before int main( void ) will make the error go away. If that's the case, it'll require a fair bit of work to solve your problem, since a tonne of source code files will need to be edited. Your best bet for getting the FFTS to work, then, would be to download gcc and install it on your sun box - I'm pretty sure it's available and know it's free. The best version of gcc to get is probably 2.96. > --- Haifa Touati <Ha...@he...> wrote: > > Hi, > > I'm installing FFTS v 0.5 on a Sun 2.8 station. > > I have installed Qt, PostgreSQL and LibCurl as mentionned in > > the > > installation guide. > > I have changed the Makefile to use CC sparc compiler instead > > of the g++ one > > (because this latter isn't very well installed on my > station). > > when I lunch the MakeAll, I receive errors like : > > type name expected instead of string. > > So I added in the MakeVars file the stl lib "string.h" > > but errors persist. > > How can I resolve this? > > Does FFTS installation requires g++? > > Thanks in advance. > > Haifa. > > > > > > _______________________________________________ > > ffts-develop mailing list > > fft...@li... > > https://lists.sourceforge.net/lists/listinfo/ffts-develop > > > ===== > ------------------------------------------------------- > Mark Higgins, mgh...@ya... > http://www.molala.net > > Opinions expressed herein do not reflect my own views; > I haven't had free will since last year when aliens ate > my brain. > > __________________________________________________ > Terrorist Attacks on U.S. - How can you help? > Donate cash, emergency relief information > http://dailynews.yahoo.com/fc/US/Emergency_Information/ > ===== ------------------------------------------------------- Mark Higgins, mgh...@ya... http://www.molala.net Opinions expressed herein do not reflect my own views; I haven't had free will since last year when aliens ate my brain. __________________________________________________ Terrorist Attacks on U.S. - How can you help? Donate cash, emergency relief information http://dailynews.yahoo.com/fc/US/Emergency_Information/ |
From: Haifa T. <Ha...@he...> - 2001-09-18 15:04:32
|
Hi, I'm installing FFTS v 0.5 on a Sun 2.8 station. I have installed Qt, PostgreSQL and LibCurl as mentionned in the installation guide. I have changed the Makefile to use CC sparc compiler instead of the g++ one (because this latter isn't very well installed on my station). when I lunch the MakeAll, I receive errors like : type name expected instead of string. So I added in the MakeVars file the stl lib "string.h" but errors persist. How can I resolve this? Does FFTS installation requires g++? Thanks in advance. Haifa. |
From: Mark H. <mgh...@ya...> - 2001-08-04 21:39:46
|
I've released version 0.7 of the Finicky Financial Trading System. Basically, this release introduces the integrated front end for the FFTS and cleans up a lot of rough edges. Here's the changelog for the new release: * GUI download tools in new Download project * Developed main front-end app FFTS (in App project) to a usable state * Evaluation graph fixes to accomodate clearing out scopes in report calcs * Added Listed Stock Option wrapper in TradeEntry to select from marked strikes and expiration dates only; changed Stock Option to OTC Stock Option to make the difference explicit. Same StockOption object underlies both wrappers, though. * Significantly reduced memory usage of evaluation graph during reporting by adding ability to delete stored data from unused scopes. * Added IndexPrice, IndexFuture, and IndexOption (very much still in dev) * Added ability to save/load quotes in TradeEntry * Added ability to process trade events (exercise/expire/etc) in TradeEntry * Fixes to make everything work properly under RedHat 7.1/gcc 2.96 * Revamped MDE to use more efficient tables and excised several bugs * Fixed PNL reports to move through market state properly and to use bilinear vol interpolation everywhere (which means it matches the market option prices exactly). * Added basic printing capability to Plotter and ReportBlotter. ===== ------------------------------------------------------- Mark Higgins, mgh...@ya... http://www.molala.net Opinions expressed herein do not reflect my own views; I haven't had free will since last year when aliens ate my brain. __________________________________________________ Do You Yahoo!? Make international calls for as low as $.04/minute with Yahoo! Messenger http://phonecard.yahoo.com/ |
From: Mark H. <mgh...@ya...> - 2001-06-01 12:06:30
|
I've just updated the cvs archive for the FFTS at http://sourceforge.net/cvs/?group_id=9699 I will now be developing using this on-line archive, so it should stay up-to-date. So you can now access the bleeding-edge code through anonymous cvs access. ===== ------------------------------------------------------- Mark Higgins, mgh...@ya... http://www.molala.net Opinions expressed herein do not reflect my own views; I haven't had free will since last year when aliens ate my brain. __________________________________________________ Do You Yahoo!? Get personalized email addresses from Yahoo! Mail - only $35 a year! http://personal.mail.yahoo.com/ |
From: Mark H. <mgh...@ya...> - 2001-05-29 01:24:36
|
I've put out a new release of the FFTS (v0.6). There are a bunch of little changes, but the main change is the way that equity options handle dividend yields. Before this release, the dividend yield for a stock was saved on the stock price object, and assumed constant across tenors. With this release, the dividend yields for the different option expiration dates are backed out from the closest-to-ATM option prices - it looks at the value of (call - put), which is a monotonic function of dividend yield even for American options. Really, to do it even more correctly, it should account for the *discrete* dividend payments, rather than assuming a continuous yield. This would be better for risk management, but it's more complex, and it's not clear how to apply the Baron-Adesi and Whaley approximation with continuous dividends. Something to think about, though. I'm also pushing the current cvs tree out to the sourceforge site (https://sourceforge.net/cvs/?group_id=9699). It should be there within a couple days (the current stuff there is pretty old). The new cvs tree contains a bunch of stuff that's new after v0.6, including some fixes to the evaluation graph. ===== ------------------------------------------------------- Mark Higgins, mgh...@ya... http://www.molala.net Opinions expressed herein do not reflect my own views; I haven't had free will since last year when aliens ate my brain. __________________________________________________ Do You Yahoo!? Yahoo! Auctions - buy the things you want at great prices http://auctions.yahoo.com/ |
From: Mark H. <mgh...@ya...> - 2001-05-20 15:56:32
|
Hmmm... that is odd. It doesn't complain with me for some reason. I'll add that include, though; probably some weird compiler thing. Let me know if you manage to get it running! --- s d <sea...@ho...> wrote: > After some more messing around, I found that the object > OptionDataFns > wouldn't compile, therefore libCommon never got built ( see > the snippet > below). Looking at OptionDataFns.cpp, it seems that function > fabs is never > defined. I added #include <cmath>. It now builds. Plan to > spend some time > playing with it. > > How did everybody else get this to build? > jay > > common... > g++ -fPIC -c CObjectFactory.cpp -L/usr/share/pgsql/ > -I/usr/include/pgsql/ > -L/opt/ffts/FFTS/lib -I/opt/ffts/FFTS/include > -ftemplate-depth-36 > g++ -fPIC -c OptionDataFns.cpp -L/usr/share/pgsql/ > -I/usr/include/pgsql/ > -L/opt/ffts/FFTS/lib -I/opt/ffts/FFTS/include > -ftemplate-depth-36 > OptionDataFns.cpp: In function `void FilterOptionQuotes > (vector<OptionQuote, allocator<OptionQuote> > &, const Date &, > double)': > OptionDataFns.cpp:88: warning: choosing > `NumericValue<T>::operator T () > [with T = double]' over `NumericValue<T>::operator const T () > const > [with T = double]' > OptionDataFns.cpp:88: warning: for conversion from `Double' > to > `double' > OptionDataFns.cpp:88: warning: because conversion sequence > for the > argument is better > OptionDataFns.cpp:106: `fabs' undeclared (first use this > function) > OptionDataFns.cpp:106: (Each undeclared identifier is reported > only > once for each function it appears in.) > OptionDataFns.cpp:122: warning: choosing > `NumericValue<T>::operator T > () [with T = double]' over `NumericValue<T>::operator const T > () const > [with T = double]' > OptionDataFns.cpp:122: warning: for conversion from `Double' > to > `double' > OptionDataFns.cpp:122: warning: because conversion sequence > for the > argument is better > make: *** [OptionDataFns.o] Error 1 > utils... > > > _________________________________________________________________ > Get your FREE download of MSN Explorer at > http://explorer.msn.com > ===== ------------------------------------------------------- Mark Higgins, mgh...@ya... http://www.molala.net Opinions expressed herein do not reflect my own views; I haven't had free will since last year when aliens ate my brain. __________________________________________________ Do You Yahoo!? Yahoo! Auctions - buy the things you want at great prices http://auctions.yahoo.com/ |
From: s d <sea...@ho...> - 2001-05-19 13:25:16
|
After some more messing around, I found that the object OptionDataFns wouldn't compile, therefore libCommon never got built ( see the snippet below). Looking at OptionDataFns.cpp, it seems that function fabs is never defined. I added #include <cmath>. It now builds. Plan to spend some time playing with it. How did everybody else get this to build? jay common... g++ -fPIC -c CObjectFactory.cpp -L/usr/share/pgsql/ -I/usr/include/pgsql/ -L/opt/ffts/FFTS/lib -I/opt/ffts/FFTS/include -ftemplate-depth-36 g++ -fPIC -c OptionDataFns.cpp -L/usr/share/pgsql/ -I/usr/include/pgsql/ -L/opt/ffts/FFTS/lib -I/opt/ffts/FFTS/include -ftemplate-depth-36 OptionDataFns.cpp: In function `void FilterOptionQuotes (vector<OptionQuote, allocator<OptionQuote> > &, const Date &, double)': OptionDataFns.cpp:88: warning: choosing `NumericValue<T>::operator T () [with T = double]' over `NumericValue<T>::operator const T () const [with T = double]' OptionDataFns.cpp:88: warning: for conversion from `Double' to `double' OptionDataFns.cpp:88: warning: because conversion sequence for the argument is better OptionDataFns.cpp:106: `fabs' undeclared (first use this function) OptionDataFns.cpp:106: (Each undeclared identifier is reported only once for each function it appears in.) OptionDataFns.cpp:122: warning: choosing `NumericValue<T>::operator T () [with T = double]' over `NumericValue<T>::operator const T () const [with T = double]' OptionDataFns.cpp:122: warning: for conversion from `Double' to `double' OptionDataFns.cpp:122: warning: because conversion sequence for the argument is better make: *** [OptionDataFns.o] Error 1 utils... _________________________________________________________________ Get your FREE download of MSN Explorer at http://explorer.msn.com |
From: s d <sea...@ho...> - 2001-05-19 02:02:19
|
Ok. Tried again. export QTDIR ,PSDIR, FFTS_DEV_DIR. source FFTS_Setup. ./MakeAll. Now it dies because it can't find lCommon. Here's the output from the make. AFAICS, it doesn't actually build libCommon.so and copy it to lib/. MakeAll tells it to build. I looked at the Makefile in common/. Can't figure out what's the problem. FWIW, I have make- 3.79.1. thanks jay . . . trade... g++ -fPIC -c CAccount.cpp -L/usr/share/pgsql/ -I/usr/include/pgsql/ -L/opt/ffts/FFTS/lib -I/opt/ffts/FFTS/include -ftemplate-depth-36 g++ -fPIC -c CTrade.cpp -L/usr/share/pgsql/ -I/usr/include/pgsql/ -L/opt/ffts/FFTS/lib -I/opt/ffts/FFTS/include -ftemplate-depth-36 g++ -fPIC -c CPositionList.cpp -L/usr/share/pgsql/ -I/usr/include/pgsql/ -L/opt/ffts/FFTS/lib -I/opt/ffts/FFTS/include -ftemplate-depth-36 g++ -fPIC -c TradeFns.cpp -L/usr/share/pgsql/ -I/usr/include/pgsql/ -L/opt/ffts/FFTS/lib -I/opt/ffts/FFTS/include -ftemplate-depth-36 g++ -fPIC -c CPortfolio.cpp -L/usr/share/pgsql/ -I/usr/include/pgsql/ -L/opt/ffts/FFTS/lib -I/opt/ffts/FFTS/include -ftemplate-depth-36 g++ -shared -Wl,-soname,libTrade.so -o libTrade.so CAccount.o CTrade.o CPositionList.o TradeFns.o CPortfolio.o cp libTrade.so /opt/ffts/FFTS/lib/libTrade.so common... g++ -fPIC -c CObjectFactory.cpp -L/usr/share/pgsql/ -I/usr/include/pgsql/ -L/opt/ffts/FFTS/lib -I/opt/ffts/FFTS/include -ftemplate-depth-36 g++ -fPIC -c OptionDataFns.cpp -L/usr/share/pgsql/ -I/usr/include/pgsql/ -L/opt/ffts/FFTS/lib -I/opt/ffts/FFTS/include -ftemplate-depth-36 utils... g++ -fPIC -c CCurveParser.cpp -L/usr/share/pgsql/ -I/usr/include/pgsql/ -L/opt/ffts/FFTS/lib -I/opt/ffts/FFTS/include -ftemplate-depth-36 g++ -fPIC -c CYahooCSVParser.cpp -L/usr/share/pgsql/ -I/usr/include/pgsql/ -L/opt/ffts/FFTS/lib -I/opt/ffts/FFTS/include -ftemplate-depth-36 >This is the source of problem (2), I'm afraid - you need to have >libCommon in there. > >How are you building the FFTS? libCommon should get built before >you need it for anything. > >You should do the following, in this order: > >1) Extract the FFTS code into some directory. I'll call it >/home/FFTS, but it can be whatever you want. > >2) Set the environment variable FFTS_DEV_DIR to the directory >you extracted the FFTS code, like > >export FFTS_DEV_DIR=/home/FFTS > >in bash. > >3) Change to $FFTS_DEV_DIR and do the following: > >source FFTS_Setup >./MakeAll > >This should build a bunch of projects, and eventually build >the "common" project. Poke around in the MakeAll script to >see what it's doing (it's just command-line bash). > _________________________________________________________________ Get your FREE download of MSN Explorer at http://explorer.msn.com |
From: Mark H. <mgh...@ya...> - 2001-05-18 03:40:18
|
> I've had some problems making 0.5. One I fixed, the other I > need help > > 1. MakeVars refers to a libCommon ( -lCommon ) which gave me > series of ld > errors because it wasn't in my tarball. That I fixed by > removing it from > MakeVars. This is the source of problem (2), I'm afraid - you need to have libCommon in there. How are you building the FFTS? libCommon should get built before you need it for anything. You should do the following, in this order: 1) Extract the FFTS code into some directory. I'll call it /home/FFTS, but it can be whatever you want. 2) Set the environment variable FFTS_DEV_DIR to the directory you extracted the FFTS code, like export FFTS_DEV_DIR=/home/FFTS in bash. 3) Change to $FFTS_DEV_DIR and do the following: source FFTS_Setup ./MakeAll This should build a bunch of projects, and eventually build the "common" project. Poke around in the MakeAll script to see what it's doing (it's just command-line bash). Really, it should use the standard configure / make / make install, but I haven't set it up yet, so it's a bit non-standard. You shouldn't have to remove anything from MakeVars; in fact, if you do, you'll just run into linking problems, since a lot of the projects are needed for the applications (ReportBlotter was the one you had in your example, I think). ===== ------------------------------------------------------- Mark Higgins, mgh...@ya... http://www.molala.net Opinions expressed herein do not reflect my own views; I haven't had free will since last year when aliens ate my brain. __________________________________________________ Do You Yahoo!? Yahoo! Auctions - buy the things you want at great prices http://auctions.yahoo.com/ |
From: s d <sea...@ho...> - 2001-05-17 03:34:28
|
I've had some problems making 0.5. One I fixed, the other I need help 1. MakeVars refers to a libCommon ( -lCommon ) which gave me series of ld errors because it wasn't in my tarball. That I fixed by removing it from MakeVars. 2. I'm getting a whole series of errors, which finally makes compilation die, on ObjectFactory:: objects. Is this related to libCommon?? I've copied below three of the error mesages. FFTS sounds like a very interesting program. I'd like to get it to work. thanks jay g++ -o ReportBlotter ReportBlotter.cpp -L/usr/lib/pgsql -I/usr/include/pgsql -L /opt/ffts/FFTS/lib -I/opt/ffts/FFTS/include -L/usr/lib/qt-2.3.0/lib -I/usr/lib/q t-2.3.0/include -ftemplate-depth-36 -lTypes -lFFTSDB -lBase -lTree -lAnalytics -lTrade -lString -lCcy -lEquity -lExotic -lpq++ -lqt -lReportBlotter -lReport /opt/ffts/FFTS/lib/libTree.so: undefined reference to `ObjectFactory::ExtractMar ketObject(BaseSession &, basic_string<char, string_char_traits<char>, __default_ alloc_template<true, 0> > const &, basic_string<char, string_char_traits<char>, __default_alloc_template<true, 0> > const &)' /opt/ffts/FFTS/lib/libTree.so: undefined reference to `ObjectFactory::ExtractPor tfolioObject(BaseSession &, basic_string<char, string_char_traits<char>, __defau lt_alloc_template<true, 0> > const &, basic_string<char, string_char_traits<char >, __default_alloc_template<true, 0> > const &)' collect2: ld returned 1 exit status make: *** [ReportBlotter] Error 1 g++ -o TradeEntry TradeEntry.cpp -L/usr/lib/pgsql -I/usr/include/pgsql -L/opt/ffts/FFTS/lib -I/opt/ffts/FFTS/include -L/usr/lib/qt-2.3.0/lib -I/usr/lib/qt-2.3.0/include -ftemplate-depth-36 -lTypes -lFFTSDB -lBase -lTree -lAnalytics -lTrade -lString -lCcy -lEquity -lExotic -lpq++ -lqt -lWidgets -lReport -lReportBlotter -lTradeEntry /opt/ffts/FFTS/lib/libTree.so: undefined reference to `ObjectFactory::ExtractMarketObject(BaseSession &, basic_string<char, string_char_traits<char>, __default_alloc_template<true, 0> > const &, basic_string<char, string_char_traits<char>, __default_alloc_template<true, 0> > const &)' /opt/ffts/FFTS/lib/libTree.so: undefined reference to `ObjectFactory::ExtractPortfolioObject(BaseSession &, basic_string<char, string_char_traits<char>, __default_alloc_template<true, 0> > const &, basic_string<char, string_char_traits<char>, __default_alloc_template<true, 0> > const &)' collect2: ld returned 1 exit status make: *** [TradeEntry] Error 1 g++ -o MDE MDE.cpp -L/usr/lib/pgsql -I/usr/include/pgsql -L/opt/ffts/FFTS/lib -I/opt/ffts/FFTS/include -L/usr/lib/qt-2.3.0/lib -I/usr/lib/qt-2.3.0/include -ftemplate-depth-36 -lTypes -lFFTSDB -lBase -lTree -lAnalytics -lTrade -lString -lCcy -lEquity -lExotic -lpq++ -lqt -lWidgets -lMDE /opt/ffts/FFTS/lib/libTree.so: undefined reference to `ObjectFactory::ExtractMarketObject(BaseSession &, basic_string<char, string_char_traits<char>, __default_alloc_template<true, 0> > const &, basic_string<char, string_char_traits<char>, __default_alloc_template<true, 0> > const &)' /opt/ffts/FFTS/lib/libTree.so: undefined reference to `ObjectFactory::ExtractPortfolioObject(BaseSession &, basic_string<char, string_char_traits<char>, __default_alloc_template<true, 0> > const &, basic_string<char, string_char_traits<char>, __default_alloc_template<true, 0> > const &)' collect2: ld returned 1 exit status make: *** [MDE] Error 1 g++ -shared -Wl,-soname,libPlotter.so -o libPlotter.so CRowBase.o CRowStockClose.o CRowFactory.o CSeriesDesc.o CPlotter.o CRowHistVol.o CRowHistVolHLOC.o CRowATMVol.o CRowOptVol.o cp libPlotter.so /opt/ffts/FFTS/lib/libPlotter.so g++ -o Plotter Plotter.cpp -L/usr/lib/pgsql -I/usr/include/pgsql -L/opt/ffts/FFTS/lib -I/opt/ffts/FFTS/include -L/usr/lib/qt-2.3.0/lib -I/usr/lib/qt-2.3.0/include -ftemplate-depth-36 -lTypes -lFFTSDB -lBase -lTree -lAnalytics -lTrade -lString -lCcy -lEquity -lExotic -lpq++ -lqt -lWidgets -lPlotter /opt/ffts/FFTS/lib/libTree.so: undefined reference to `ObjectFactory::ExtractMarketObject(BaseSession &, basic_string<char, string_char_traits<char>, __default_alloc_template<true, 0> > const &, basic_string<char, string_char_traits<char>, __default_alloc_template<true, 0> > const &)' /opt/ffts/FFTS/lib/libTree.so: undefined reference to `ObjectFactory::ExtractPortfolioObject(BaseSession &, basic_string<char, string_char_traits<char>, __default_alloc_template<true, 0> > const &, basic_string<char, string_char_traits<char>, __default_alloc_template<true, 0> > const &)' /opt/ffts/FFTS/lib/libPlotter.so: undefined reference to `FilterOptionQuotes(vector<OptionQuote, allocator<OptionQuote> > &, Date const &, double)' collect2: ld returned 1 exit status make: *** [Plotter] Error 1 _________________________________________________________________ Get your FREE download of MSN Explorer at http://explorer.msn.com |
From: Mark H. <mgh...@ya...> - 2001-04-13 19:06:52
|
I have released v. 0.5 of the FFTS today. I completely revamped the evaluation graph code for this release, so it's a pretty big change. The old graph stuff was unstable and would try to read deleted memory, etc. The new code is much more stable and more efficient. The projects have been reorganized a bit as well; I stripped all of the equity-specific code out into its own project, to make it a bit easier to extend the FFTS to other markets. Another big change: extensive regression tests. Many of the projects now have a tests subdirectory, which contains a bunch of unit tests. As problems appear and are fixed, new cases can be added to the test suite to catch them should they arise again. Note that the cvs repository hasn't been updated yet, so you'll have to download the tarball. You can get it through http://ffts.sourceforge.net ===== ------------------------------------------------------- Mark Higgins, mgh...@ya... http://www.molala.net Opinions expressed herein do not reflect my own views; I haven't had free will since last year when aliens ate my brain. __________________________________________________ Do You Yahoo!? Get email at your own domain with Yahoo! Mail. http://personal.mail.yahoo.com/ |
From: Paul G. <gr...@me...> - 2001-03-14 03:25:49
|
I'll have a play around with this code, to see what's wrong ... just seems like the format has changed slightly. I have a couple of web-servers one in the US and one in HK so perhaps I can get them running automatically to produce the files, i.e. a complete DB and a daily update ... would this be useful to you? One of my programmers is already working on some code to download the HK data into a Sybase database, we use python for this stuff. I can try to put a database online and give public access to it. Sybase is pretty lightweight and free for development purposes on Linux. I think this is the most flexible way. <SNIP> 3) I've sort of stopped gathering daily market data for stocks and options (I had to cut off my lovely always-on internet connection a while ago :( ), but I did notice that the program I use to snap and save option prices wasn't working right. Not sure what's up with that; the work on the fundamental architecture hasn't given me much time to play with the higher level stuff. </SNIP> |
From: Mark H. <mgh...@ya...> - 2001-03-14 01:46:11
|
A few points: 1) I've been spending the last couple of months basically gutting the underlying evaluation graph stuff and rebuilding it to make it more stable. There are a lot of changes in the code that aren't even in cvs on soureforge... though they should be within a couple of weeks. 2) Odd lots shouldn't be a problem. Right now, the option object is actually *more* suited to this than to an exchange-traded option - it really corresponds to an option on a single stock, so you can do as many units as you like. And the expiration date is saved just as a date, not as a pointer to a listed option expiration month or something. I suspect that I'll keep this flexible option object around (which you could easily use for your warrants), and build a new "listed option" object to make it easier to do normal exchange-based options trading. Building new classes to represent different types of security isn't completely straightforward, but it's not too hard either, especially if the new security types can use market data in common with other types (so you don't need to build classes for the market data as well). I've tried to design the ffts so that it's relatively easy to extend and customize this way. 3) I've sort of stopped gathering daily market data for stocks and options (I had to cut off my lovely always-on internet connection a while ago :( ), but I did notice that the program I use to snap and save option prices wasn't working right. Not sure what's up with that; the work on the fundamental architecture hasn't given me much time to play with the higher level stuff. Anyway, watch this space for more info on the next version of the ffts, which should be appearing within the month. > Message: 1 > From: "Paul Gresham" <gr...@me...> > To: <fft...@li...> > Date: Tue, 13 Mar 2001 21:55:19 +0800 > Subject: [Ffts-develop] Hi ... Lot sizes > > Hi Mark, > I can't believe an EQD system open source ... anyway it's a > great effort, I > had to make some changes to remove the ^M and replace them > with 013's in one > of the web grabber files ... (I'll send more info later on) > ... Anyhow it > built although I've had a few problems downloading the stock > price data to > test it out. > > We are a small development company and have built a website > for Hong Kong > (www.quotehk.com user:demo pass:demo) It's not really live yet > and certainly > doesn't have live prices. We are developing some EQD stuff for > it, we are a > bit late to use FFTS underneath it all, but perhaps can use > some or all of > the FFTS stuff in the future. > > We came upon a bit of a problem with the FFTS when we tried to > set it up for > HK Stock options and Warrants ( the warrants here are mostly > issued by third > parties and are basically options with funny expiries ). We > need to modify > FFTS to allow different lot sizes, i.e. the lot size of the > stock varies in > HK, and the number of stocks per option contract is the lot > size, it can > vary from 400 to 2000 shares per contract. Do you think this > would be an > easy mod to make, and do you have any pointers for me as to > what would need > to be modified. I did do some searching although not > extensively so I'm > being a bit lazy in asking for this! > > All the best > Paul > > > > > --__--__-- > > _______________________________________________ > ffts-develop mailing list > fft...@li... > http://lists.sourceforge.net/lists/listinfo/ffts-develop > > > End of ffts-develop Digest ===== ------------------------------------------------------- Mark Higgins, mgh...@ya... http://www.molala.net Opinions expressed herein do not reflect my own views; I haven't had free will since last year when aliens ate my brain. __________________________________________________ Do You Yahoo!? Yahoo! Auctions - Buy the things you want at great prices. http://auctions.yahoo.com/ |
From: Paul G. <gr...@me...> - 2001-03-13 14:01:40
|
Hi Mark, I can't believe an EQD system open source ... anyway it's a great effort, I had to make some changes to remove the ^M and replace them with 013's in one of the web grabber files ... (I'll send more info later on) ... Anyhow it built although I've had a few problems downloading the stock price data to test it out. We are a small development company and have built a website for Hong Kong (www.quotehk.com user:demo pass:demo) It's not really live yet and certainly doesn't have live prices. We are developing some EQD stuff for it, we are a bit late to use FFTS underneath it all, but perhaps can use some or all of the FFTS stuff in the future. We came upon a bit of a problem with the FFTS when we tried to set it up for HK Stock options and Warrants ( the warrants here are mostly issued by third parties and are basically options with funny expiries ). We need to modify FFTS to allow different lot sizes, i.e. the lot size of the stock varies in HK, and the number of stocks per option contract is the lot size, it can vary from 400 to 2000 shares per contract. Do you think this would be an easy mod to make, and do you have any pointers for me as to what would need to be modified. I did do some searching although not extensively so I'm being a bit lazy in asking for this! All the best Paul |
From: Mark H. <mgh...@ya...> - 2000-10-29 15:24:03
|
Version 0.4 of the Finicky Financial Trading System has just been released. http://ffts.sourceforge.net/Download.html The CVS repository for the project is now on-line; you can get to it through the project's SourceForge page at http://sourceforge.net/projects/ffts/ I added the Plotter, an app which plots market data - closing stock prices, historical volatilities, etc. It is set up to make adding new types of time series quite easy. Right now it includes series for closing stock prices, historical vols calculated from close to close, and historical vols calculated from high/low/open/close data (which is about 3X as accurate in terms of standard deviation around a "true" value). Using Dirk Eddelbuettel's suggestions, the make system was generalised to make compiling with libraries and headers in different directories work. The report system was also generalised a fair bit, and lots of other cleanup was done. ===== ------------------------------------------------------- Mark Higgins, mgh...@ya... http://www.molala.net Opinions expressed herein do not reflect my own views; I haven't had free will since last year when aliens ate my brain. __________________________________________________ Do You Yahoo!? Yahoo! Messenger - Talk while you surf! It's FREE. http://im.yahoo.com/ |