Timing[] and AbsoluteTime[].
- To: mathgroup at smc.vnet.net
- Subject: [mg20257] Timing[] and AbsoluteTime[].
- From: Guilherme Roschke <gr at network3.entropy.upenn.edu>
- Date: Sun, 10 Oct 1999 00:04:08 -0400
- Organization: University of Pennsylvania
- Sender: owner-wri-mathgroup at wolfram.com
How does timing measure its "value"?
here is my problem. I have fuction f[x], which does a few hundred
NIntegrates.
Timing[f[x];]
= {1770.60,Null}.
However:
before=AbsoluteTime[]; (# of secs since 1/1/1900)
Timing[f[x];]
= {1770.60,Null}
after=AbsoluteTime[];
Diff=after-before
= 14,673.000
Timing is off by a factor of 10!
Note that this is on a DEC with 2 21264 500mhz Alpha chips, where nothing
else is running except for the system deamons (lpd, sendmail, xserver
etc...).
This is especially disconcerting because the AbsoluteTime[] elapsed is
closer to a PII350 that I have than I would have hoped for in doing
numerics on the Alpha machine.
Does anyone have any thoughts of what might be going on? On how I could
actually receive my answer in 1,700 rather than 14,000 seconds?
I'm using version 4 for Digital Unix.
thanks,
Guilherme
*************************************************
Guilherme Roschke *
Programmer/SysAdmin/Researcher *
Dept. of Anesthesia *
University of Pennsylvania Medical School *
gr at network3.entropy.upenn.edu *
*************************************************
- Follow-Ups:
- Re: Timing[] and AbsoluteTime[].
- From: Jean-Marie Thomas <jmt@agat.net>
- Re: Timing[] and AbsoluteTime[].