sort

[Viper]

i had heard there was a build in sorter in DevC++ is this true?? if so could someone show me how to use it plz n thx

[wtd]

Dev-C++ is an environment for writing C++ code, not a language.

C++'s Standard Template Library does have a "sort" function.

http://www.fredosaurus.com/notes-cpp/algorithms/sorting/stl-sort-arrays.html

[[Gandalf]]

Is qsort() (quick sort) part of the same library, the math one, or some other?

[wtd]

code:

#include <cstdlib>

Don't use qsort.

[Viper]

what is tht include for?? n thx for the help with the sort

[wtd]

It's a C++ wrapper for C's standard library.

[[Gandalf]]

wtd wrote:

code:

#include <cstdlib>

Don't use qsort.

Why not?

[wtd]

Bedcause qsort resorts to all kinds of void pointer and function pointer madness. The STL's sort function uses templates and operator overloading to accomplish the same thing in a type-safe, convenient way.

[bugzpodder]

qsort is a pain in the ass to use. stick with sort and you'll be fine.

[md]

Of course you can always write your own sort function; that way you better understand how it works... and you can customize the sort to your particular needs... a insert sort is faster then a quick sort with mostly sorted data

[wtd]

The STL sort function is quite fast for most cases.

[bugzpodder]

Cornflake wrote:

Of course you can always write your own sort function; that way you better understand how it works... and you can customize the sort to your particular needs... a insert sort is faster then a quick sort with mostly sorted data

that is, assuming quicksort uses the first element as the pivot. I can assure you that any reasonable programmer would use a random pivot or maybe even median of three to implement pivot function in quicksort. in this case, quicksort would perform better. However, it is worthwhile to mention that selection sort (not insertion sort) performs better on smaller datasets (maybe around 20) than quicksort, so a decent quicksort implementation would switch to selection sort when the partition is less than 20. so rest assured that most built-in sorting algorithms are near optimal.