C++Talk.NET C++ language newsgroups   Archives   FAQ   Search   Memberlist   Usergroups   Register   Profile   Log in to check your private messages   Log in  Must second bind2nd parameter be constant for loop?          C++Talk.NET Forum Index -> C++ Language (Moderated) View previous topic :: View next topic   Author Message Rex Kerr Guest Posted: Sat Dec 09, 2006 10:10 am    Post subject: Must second bind2nd parameter be constant for loop? I'm trying to compare two vectors to ensure that their values are equal within a tollerance (easy, std::equal with a binary predicate), but my definition of equal also allows for a certain number of values to lie outside of the tollerance, so I want something like std::count_if that takes two ranges and passes them to a binary op. So, I've tried using bind2nd and using a post-incremented iterator as the second parameter (which doesn't work), but then I also realized that I'm making an assumption that the count_if algorithm works sequentially from begin to end, which it may not. So, how would I do such a thing using standard algorithms? Here's what I have that doesn't work: namespace { template <class T> struct OutsideTollerance : public std::binary_function<T, T, T> { T operator() (T v1, T v2) const { return std::abs(v2 - v1) >= 2; // hardcoded for example } }; } std::size_t num_mismatched = std::count_if( v1.begin(), v1.end(), std::bind2nd(OutsideTollerance<BYTE>(), *(v2it++)) ); (variable names changed to protect the innocent) Thanks! -- [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Guest Posted: Sun Dec 10, 2006 9:29 am    Post subject: Re: Must second bind2nd parameter be constant for loop? I would try this, although I understand you might not like using const_cast. struct EuqalWithTollerance : public std::binary_function<double, double, bool> { EuqalWithTollerance( double tol, int maxOutOfRange ) : tol_(tol), mOut_(maxOutOfRange), nOut_(0) {}; bool operator() (double v1, double v2) const { return (fabs(v2 - v1) < tol_)? true : (++const_cast<int&>(nOut_) < mOut_); } double tol_; int nOut_; int mOut_; }; std::equal( v1.begin(), v1.end(), v2.begin(), EuqalWithTollerance( 1e-4, 3 ) ); Rex Kerr wrote: Quote: I'm trying to compare two vectors to ensure that their values are equal within a tollerance (easy, std::equal with a binary predicate), but my definition of equal also allows for a certain number of values to lie outside of the tollerance, so I want something like std::count_if that takes two ranges and passes them to a binary op. So, I've tried using bind2nd and using a post-incremented iterator as the second parameter (which doesn't work), but then I also realized that I'm making an assumption that the count_if algorithm works sequentially from begin to end, which it may not. So, how would I do such a thing using standard algorithms? Here's what I have that doesn't work: namespace { template <class T struct OutsideTollerance : public std::binary_function<T, T, T { T operator() (T v1, T v2) const { return std::abs(v2 - v1) >= 2; // hardcoded for example } }; } std::size_t num_mismatched = std::count_if( v1.begin(), v1.end(), std::bind2nd(OutsideTollerance<BYTE>(), *(v2it++)) ); (variable names changed to protect the innocent) Thanks! -- [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Mathias Gaunard Guest Posted: Sun Dec 10, 2006 10:10 am    Post subject: Re: Must second bind2nd parameter be constant for loop? Rex Kerr wrote: Quote: Here's what I have that doesn't work: namespace { template <class T struct OutsideTollerance : public std::binary_function<T, T, T { T operator() (T v1, T v2) const { return std::abs(v2 - v1) >= 2; // hardcoded for example } }; } std::size_t num_mismatched = std::count_if( v1.begin(), v1.end(), std::bind2nd(OutsideTollerance<BYTE>(), *(v2it++)) ); What's `v2it'? -- [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Pete Becker Guest Posted: Sun Dec 10, 2006 10:10 am    Post subject: Re: Must second bind2nd parameter be constant for loop? Rex Kerr wrote: Quote: I'm trying to compare two vectors to ensure that their values are equal within a tollerance (easy, std::equal with a binary predicate), but my definition of equal also allows for a certain number of values to lie outside of the tollerance, so I want something like std::count_if that takes two ranges and passes them to a binary op. count_if won't help here, because it doesn't take two ranges. If you insist on using algorithms from the standard library, I think you've got a two step process. First, use transform to walk through the two ranges and write out a sequence giving the result of each comparison of pairs of values. Then walk through that sequence and count the number of values that are out of tolerance. If you want to roll your own algorithm, it's straightforward: template <class InIt1, class InIt2> int out_of_tolerance(InIt1 first1, InIt1 last1, InIt2 first2) { int count = 0; while (first1 != last1) if (compare(*first1++, *first2++)) ++count; return count; } -- -- Pete Roundhouse Consulting, Ltd. (www.versatilecoding.com) Author of "The Standard C++ Library Extensions: a Tutorial and Reference." (www.petebecker.com/tr1book) [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Carlos Moreno Guest Posted: Sun Dec 10, 2006 10:10 am    Post subject: Re: Must second bind2nd parameter be constant for loop? Rex Kerr wrote: Quote: So, I've tried using bind2nd and using a post-incremented iterator as the second parameter (which doesn't work), but then I also realized that I'm making an assumption that the count_if algorithm works sequentially from begin to end, which it may not. What?? How exactly would it not go sequentially from begin to end??? That is precisely how it's described in its documented functionality. Quote: std::size_t num_mismatched = std::count_if( v1.begin(), v1.end(), std::bind2nd(OutsideTollerance<BYTE>(), *(v2it++)) ); There's an clear detail that you have to realize: count_if is a function; and bind2nd is one of its parameters!! The value of the parameter is evaluated and passed to the function --- how many times is the parameter evaluated? Well, the function is called *once*, so the parameter is evaluated that same one time, then passed to the function. I guess you would solve it with your own custom function object that keeps track (via a data member) of how many elements have fallen outside tolerance --- as soon as you exceed the limit, then the functor starts returning false, at which point std::equal will immediately return false. HTH, Carlos -- [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Seungbeom Kim Guest Posted: Mon Dec 11, 2006 7:46 am    Post subject: Re: Must second bind2nd parameter be constant for loop? fcannizzo.mifft2001 (AT) london (DOT) edu wrote: Quote: I would try this, although I understand you might not like using const_cast. struct EuqalWithTollerance : public std::binary_function<double, double, bool { EuqalWithTollerance( double tol, int maxOutOfRange ) : tol_(tol), mOut_(maxOutOfRange), nOut_(0) {}; bool operator() (double v1, double v2) const { return (fabs(v2 - v1) < tol_)? true : (++const_cast<int&>(nOut_) mOut_); } double tol_; int nOut_; int mOut_; }; std::equal( v1.begin(), v1.end(), v2.begin(), EuqalWithTollerance( 1e-4, 3 ) ); It's a nice idea, but why do you need const_cast? You can make operator() a non-const member function and get rid of const_cast. And it turns that you can save an extra variable: :) template <typename T> class EqualWithTolerance : public std::binary_function<T, T, bool> { public: EqualWithTolerance(T tol, int max_out_tol) : tol_(tol), n_(max_out_tol) { } bool operator()(double v1, double v2) { return (std::abs(v2 - v1) < tol_) || (--n_ >= 0); } private: T tol_; int n_; }; bool equal = std::equal(v1.begin(), v1.end(), v2.begin(), EqualWithTolerance(1e-4, 3)); -- Seungbeom Kim [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Matthew T. Staebler Guest Posted: Tue Dec 12, 2006 7:52 am    Post subject: Re: Must second bind2nd parameter be constant for loop? Carlos Moreno wrote: Quote: James Kanze wrote: Note too that the predicate must be copiable, and the algorithm can copy it as many times and whenever it wants. So, you're suggesting that the implementation could be recursive? I would hate to put words into James's mouth, but I assume that he was cautioning someone using algorithms such as count_if and equal from relying on predicates that store values that change as elements of the range are visited. In the EqualWithTolerance class defined above, there is an int nOut_ member whose value should be based upon the results of previous visits to other elements of the range. However, it is not guaranteed that the EqualWithTolerance object that visits one element will be the same as the EqualWithTolerance object that visits another element, as the implementation of the algorithm has the flexibility to copy the predicate object as it sees fit. It would be a perfectly legal implementation of the algorithm to create a unique copy of the passed predicate object for every element of the range. Using such an implementation with the EqualWithTolerance class as a predicate would mean that every element would be visited with an EqualWithTolerance object whose nOut_ member had a value of 0. While it is probably unlikely (although still legal) for an implementation of count_if and equal to copy the predicate object in such a fashion, it is much more likely for other algorithms such a find_if to copy the predicate object. If one is in the habit of ignoring the possibility for unlikely algorithms such as count_if, it is much easier to forget about the possibility for likely algorithms such as find_if. One fairly simple solution to this problem is to use a reference for the mutable values in your predicate object. Changing the nOut_ member in EqualWithTolerance from an int to an int& and supplying an int& to the constructor of EqualWithTolerance will resolve the issue. Matthew T. Staebler -- [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Maxim Yegorushkin Guest Posted: Wed Dec 13, 2006 5:42 am    Post subject: Re: Must second bind2nd parameter be constant for loop? Michiel.Salters (AT) tomtom (DOT) com wrote: Quote: Rex Kerr wrote: I'm trying to compare two vectors to ensure that their values are equal within a tollerance (easy, std::equal with a binary predicate), Here's what I have that doesn't work: namespace { template <class T struct OutsideTollerance : public std::binary_function<T, T, T { T operator() (T v1, T v2) const { return std::abs(v2 - v1) >= 2; // hardcoded for example } }; } std::size_t num_mismatched = std::count_if( v1.begin(), v1.end(), std::bind2nd(OutsideTollerance<BYTE>(), *(v2it++)) ); (variable names changed to protect the innocent) Let's guess - originally it was OutsideTolerance ? ;) Seriously - when using the STL, try to write proper functors. For any kind of "equal" operation, make sure that equal(a,b) and equal(a,c) implies equal(b,c). OutsideTolerance is not such an equal function. Now, your requirements will say you need OutsideTolerance to behave like this. That's life, it just means that using the STL here won't be trivial. Does count_if require a functor with properties of equal()? I guess, it's an unary functor in question... -- [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Carlos Moreno Guest Posted: Wed Dec 13, 2006 5:43 am    Post subject: Re: Must second bind2nd parameter be constant for loop? James Kanze wrote: Quote: What?? How exactly would it not go sequentially from begin to end??? By visiting the elements in some other order. I was going to start my reply with a "James, why are you being silly?" :-) What's silly about [...] Notice the --- and this time I'll emphasize it --- *I was going to* start with .... :-) I interpreted the original comment of going sequentially from begin to end in a different way --- notice that the problem in the OP was that he was somehow magically expecting his parameter to be re- evaluated for each iteration *inside the algorithm*, and somewhat he associated that idea with that of "iterating sequentially from begin to end" --- which, of course, is the obvious and straightforward way of implementing count_if. That's why your answer of "by visiting the elements in some other order" sounded to me like you were just nitpicking with something silly that didn't even make sense --- it, of course, made sense as soon as I read the rest of your message (perhaps I should have added that detail after the initial comment ) Quote: That is precisely how it's described in its documented functionality. Where is that documented. I was expecting to see in the preconditions that it accepts an input forward iterator What does the type of iterator have to do with anything? It has to do with my confused brain :-) I was overlooking the fact that by putting lower requirements in the parameters you're *not* preventing implementations that take advantage of other iterators that meet *and exceed* those requirements; IOW, my confusion was that I was expecting that if the expected iterator is an input *forward* iterator, *the* implementation must be such that it will work with that form of iterator (and therefore, it would *have to* visit the elements sequentially *starting at begin*) --- what it really requires is that *at least one* available implementation does, as you pointed out. Quote: So, you're suggesting that the implementation could be recursive? ;-) There's nothing to forbid it. The rest was, of course, mostly kidding :-) Quote: int count_if (Iterator begin, Iterator end, Function predicate) { if (begin == end) { return 0; } else { return count_if (++Iterator(begin), end) + (pred(*begin) ? 1 : 0); } } You'd have to use meta-programming techniques to specialize the algorithm---your code here won't work for InputIterators. Other than that, it's perfectly legal. I see that you also overlooked the couple of details that I overlooked --- the call to count_if should receive the parameter predicate, and pred(*begin) should have been predicate(*begin). But putting that aside, and at the risk of continuing to make a fool of myself, I have to ask: why doesn't the above work for input iterators?? They support copy-construction, and they most definitely support pre-increment (which necessarily returns a reference to the incremented value). So, what am I missing now? Carlos -- -- [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Matthew T. Staebler Guest Posted: Wed Dec 13, 2006 10:10 am    Post subject: Re: Must second bind2nd parameter be constant for loop? Carlos Moreno wrote: Quote: int count_if (Iterator begin, Iterator end, Function predicate) { if (begin == end) { return 0; } else { return count_if (++Iterator(begin), end) + (pred(*begin) ? 1 : 0); } } James Kanze wrote: You'd have to use meta-programming techniques to specialize the algorithm---your code here won't work for InputIterators. Other than that, it's perfectly legal. I see that you also overlooked the couple of details that I overlooked --- the call to count_if should receive the parameter predicate, and pred(*begin) should have been predicate(*begin). But putting that aside, and at the risk of continuing to make a fool of myself, I have to ask: why doesn't the above work for input iterators?? They support copy-construction, and they most definitely support pre-increment (which necessarily returns a reference to the incremented value). So, what am I missing now? Your implementation of count_if relies on an order of evalution that is not guaranteed by the standard. Specifically, the dereference of the begin iterator must be evaluated before the increment of the iterator copied from the begin iterator. InputIterators do not guarantee that any copies of the iterator can be dereferenced after the iterator has been incremented. There is no way to safely traverse a range using InputIterators in a manner other than sequentially. ----- Matthew T. Staebler -- [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Carlos Moreno Guest Posted: Thu Dec 14, 2006 5:48 am    Post subject: Re: Must second bind2nd parameter be constant for loop? James Kanze wrote: Quote: But putting that aside, and at the risk of continuing to make a fool of myself, I have to ask: why doesn't the above work for input iterators?? They support copy-construction, and they most definitely support pre-increment (which necessarily returns a reference to the incremented value). So, what am I missing now? [...] Note that last post-condition. The ++ on the *copy* of begin means that begin itself is no longer required to be dereferenceable, which means that (*begin) is illegal. And since order of evaluation isn't specified, a compiler could do the incrementation before the dereferencing. *sigh* ... Of course, after seeing the explanation, I can only say DUH!! How could I forget about the (rather obvious) possibility of calling count_if as: count_if (istream_iterator<int>(file), istream_iterator<int>(), ...) Thanks (and thanks Matthew) for having the patience to remind me this detail! Cheers, Carlos -- -- [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Display posts from previous: All Posts1 Day7 Days2 Weeks1 Month3 Months6 Months1 Year Oldest FirstNewest First         C++Talk.NET Forum Index -> C++ Language (Moderated) All times are GMT Page 1 of 1   Jump to: Select a forum C/C++----------------C LanguageC Language (Moderated)C++ Language (Moderated)C++ language (comp.lang.c++)C++ language, library and standards Non-English C/C++ Newsgroups----------------C++ (German)C++ (French)  You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum Powered by phpBB © 2001, 2006 phpBB Group SEO toolkit © 2004-2006 webmedic.