comp.lang.idl-pvwave archive
Messages from Usenet group comp.lang.idl-pvwave, compiled by Paulo Penteado

Home » Public Forums » archive » float function unexpectedly slow
Show: Today's Messages :: Show Polls :: Message Navigator
E-mail to friend 
Return to the default flat view Create a new topic Submit Reply
Re: float function unexpectedly slow [message #88029 is a reply to message #88028] Wed, 12 March 2014 20:50 Go to previous messageGo to previous message
chris_torrence@NOSPAM is currently offline  chris_torrence@NOSPAM
Messages: 528
Registered: March 2007
Senior Member
Hi Tim,

I'm seeing the same thing that Craig sees - on my older MacBook Pro, running IDL 8.3, IDL is about 15% faster than python 2.7.6.

Speaking as someone who knows the guts of the float function, I wouldn't bother filing a bug report. The float routine (and all the other conversion routines) are just about as fast as they are going to get. This is all "Dave Stern" code, which means it is dense but super efficient. There is just a fair amount of code involved in checking the input arguments, parsing the string & looking for decimal points, exponents, etc., and then finally creating the IDL_VARIABLE and filling in the value.

I think your time would be better spent in eliminating any loops in your programs. IDL's thread pool doesn't even begin to work until you have more than 100,000 elements in your input array, so you won't see any significant boost with multiple cores until you can pass in all of your values at once.

Hope this helps.

Cheers,
Chris
ExelisVIS
[Message index]
 
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Previous Topic: stregex fails to match underscore
Next Topic: XML to Configure IDL Programs?

-=] Back to Top [=-
[ Syndicate this forum (XML) ] [ RSS ] [ PDF ]

Current Time: Fri Oct 10 11:05:24 PDT 2025

Total time taken to generate the page: 1.11922 seconds