Pass a back_inserter as the output iterator in remove_copy_if algorithm

[silverburgh.meryl]

i have the following code, but it does not compile because I pass in
back_inserter(b ) as the output iterator in the remove_copy_if
algorithm.
My questions are:
1. why it does not compile? Passing a back_inserter() works for copy()
algorithm, why not remove_copy_if( )?
2. is there a work-around for my problem? I think one solution is to
allocate 'b' as big as 'a', but that seems to be a waste of memory for
the case of b is a lot smaller than a.

Thank you.

class EvenValue {
public:
bool operator() (float value) {
return (int)value %2 ? false : true;
}
};

float func1(vector<fl oat>& a, float max ) {

vector<float> b;
vector<float>:: iterator itr = remove_copy_if (a.begin(), a.end(),
back_inserter(b ), EvenValue() );

// do something with 'b'
return 0.0;
}

[silverburgh.meryl]

Here is the compilation error:

../A.cpp: In member function 'float func1(vector<fl oat>&()':
.../A.cpp:361: error: conversion from
'std::back_inse rt_iterator<std ::vector<float, std::allocator< float> >

' to non-scalar type '__gnu_cxx::__n ormal_iterator< float*, std::vector<flo at, std::allocator< float> > >' requested

[Andrew Koenig]

<si************ ***@gmail.com> wrote in message
news:11******** **************@ o13g2000cwo.goo glegroups.com.. .

i have the following code, but it does not compile because I pass in
back_inserter(b ) as the output iterator in the remove_copy_if
algorithm.

That's not the reason. The reason is that if you pass a back_inserter, you
get back a back_inserter as the result; you're trying to use that
back_inserter to initialize an ordinary iterator.

[silverburgh.meryl]

Thanks. Sorry for my oversight.