STL Book
in [C++]
|
From: dominic.connor on 28 Aug 2006 09:00 { obvious answers apply: FAQ, newsgroup archives, www.accu.org, please try not to repeat them. thanks. -mod } I'm running a C++ for quants course soon, and it occurs to me that I haven't got a book recommendation for STL. The target audience is people who've bumped into vector, and have some schematic knowledge of templates, but are not experts. (else they'd be training me). I have some old STL books, and some that deal with slightly more advanced issues, but have yet to see a tutorial book that I actually like. I also review C++ books for The Register (www.theregister.co.uk) (someone has to), so if there is such a thing, I'll write it up for a larger audience. DCFC the pimp [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ]
From: Thomas Tutone on 28 Aug 2006 12:22 dominic.connor(a)gmail.com wrote: > I'm running a C++ for quants course soon, and it occurs to me that I > haven't got a book recommendation for STL. The target audience is > people who've bumped into vector, and have some schematic knowledge of > templates, but are not experts. (else they'd be training me). > > I have some old STL books, and some that deal with slightly more > advanced issues, but have yet to see a tutorial book that I actually > like. > I also review C++ books for The Register (www.theregister.co.uk) > (someone has to), so if there is such a thing, I'll write it up for a > larger audience. The obvious answer is The C++ Standard Library: A Tutorial and Reference by Nicolai M. Josuttis. It is the benchmark against which all others need to be judged. There are other decent books, like Scott Meyers' Effective STL, but they are not a substitute for the foundation provided by Josuttis' book. Best regards, Tom [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ]
From: ikinal on 28 Aug 2006 18:03 Thomas Tutone wrote: > dominic.connor(a)gmail.com wrote: > > > I'm running a C++ for quants course soon, and it occurs to me that I > > haven't got a book recommendation for STL. The target audience is > > people who've bumped into vector, and have some schematic knowledge of > > templates, but are not experts. (else they'd be training me). > ... > > The obvious answer is The C++ Standard Library: A Tutorial and > Reference by Nicolai M. Josuttis... >Scott> Meyers' Effective STL, Besides Josuttis {irreplaceable} and Meyers {very interesting} there's a NEW book out that is perhaps a bit less intimidating but still helpful. "C++ Standard Library Practical Tips" by Greg Reese. I like the style, and find it informative -- I've managed to learn from it. I think it's level is what Dominic is looking for. Good amount of code listings as well. Ihor Kinal [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ]
From: pookiebearbottom on 31 Aug 2006 15:21 dominic.connor(a)gmail.com wrote: > I'm running a C++ for quants course soon, be sure to teach them that their main job is to get the numbers correct!! Just last week I saw two VERY smart Interest Rate quants arguing about the underlying represenation of the their date structure should be. When they finally asked my opinion, my answer was: Does it matter as long as it works for you and it is fast enough? Let me worry about this, and you do the "important stuff" (feeding the egos) [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ]
From: dominic.connor on 1 Sep 2006 08:50
> be sure to teach them that their main job is to get the numbers > correct!! It's a good point. Most quants have *no* CompSci background at all, and often refer to programming as "trivial". Few programming texts talk much about defensive programming, stuff like asserting values within sanity bounds, or that the worst sort of bug is one that gives you plausible, but wrong numbers. They know that an American option price is always >= European, but clearly think I'm weird when I suggest you put this stuff in as checks. > Just last week I saw two VERY smart Interest Rate quants arguing about > the underlying represenation of the their date structure should be. Oh, dear, I wonder if that's my fault... Part of what I try to teach is looking at objects at multiple levels, and I use the double/date as an example. An extra hassle of no CS knowledge is that >75% of students turn up for a C++ course not knowing what a stack is, or that a double is 8 bytes, and bits cause a few vague looks as well. 50% of students have numerate PhDs, all have BSc/MSc in a math based subject. Yet explaining ** takes a surprising amount of time. I had a "classical" training in programming where of course one tried not to worry about low level representation, but a lot of people out there aren't disregarding low-level issues, they are ignorant of them, which is a different thing altogether. It's so bad that I've had to start knocking up a rather rambling "Computer Science For Dummies" document. Please, someone tell me that there is a book already like this ? DominiConnor DCFC The Pimp [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] |