timing statistics for stl containers

[CMOS]

hi,
does any one know a place to get performance statistics for stl
containers?

CMOS

[DreamAndLove]

On 5ÔÂ21ÈÕ, ÏÂÎç3ʱ39·Ö, CMOS <manu...@millenniumit..comwrote:

hi,
does any one know a place to get performance statistics for stl
containers?

CMOS

I don't know!

[Ian Collins]

CMOS wrote:

hi,
does any one know a place to get performance statistics for stl
containers?

Your profiler.

--
Ian Collins.

[=?iso-8859-1?q?Erik_Wikstr=F6m?=]

On 21 Maj, 09:39, CMOS <[e-mail]:

hi,
does any one know a place to get performance statistics for stl
containers?

What for? Synthetic benchmarks are often close to useless in realworld
applications (except for the most trivial cases). Select the
containers you need based on the functionality they offer, if you then
get problems with performance use a profiler to locate the bottlenecks
and then start figuring out how to improve.

If you are clever and know the usage-patterns of your code there's
always a possibility to write smart code that maximizes performance
but often it will not give any noticeable difference for the user and
in return it will usually give you unmaintainable code.

--
Erik Wikström

[Mark P]

CMOS wrote:

hi,
does any one know a place to get performance statistics for stl
containers?

CMOS

The standard specifies complexity guarantees on many of the non-trivial
operations. You can usually browse the library implementations if you
locate the correct include directories on your system. If you need real
world numbers measured in seconds, you'll almost certainly have to
measure these yourself.