Re: Integrate got slower in Version 3.0/Example
- To: mathgroup at smc.vnet.net
- Subject: [mg9006] Re: Integrate got slower in Version 3.0/Example
- From: carlos at mars.Colorado.EDU (Carlos A. Felippa)
- Date: Tue, 7 Oct 1997 03:35:55 -0400
- Organization: University of Colorado, Boulder
- Sender: owner-wri-mathgroup at wolfram.com
In article <616vj5$op0 at smc.vnet.net> Elias Saab <mathes at math.missouri.edu> writes:
>
>In version 2.2
>
>In[1]:= Integrate[Sqrt[R^2-x^2-y^2],{x,-R,R},
> {y,-Sqrt[R^2-x^2],Sqrt[R^2-x^2]}]//Timing
>
>
> 3
> 2 Pi R
>Out[1]= {5.17 Second, -------}
> 3
>
>In Version 3.0
>Integrate[Sqrt[R^2-x^2-y^2],{x,-R,R},
> {y,-Sqrt[R^2-x^2],Sqrt[R^2-x^2]}]//Timing
>Integrate::"gener": "Unable to check convergence"
>
>
> 3
> 2 Pi R
>Out[1]= {32.8 Second, -------}
> 3
>
>I mentionned this to Wolfram before the release of 3.0. It seems
>that nothing has been done to correct this.
>
>Elias Saab
>
>
I noticed that too on my Mac versions. For the above integrals:
Mma 2.2: 4.7667 Seconds
Mma 3.0: 22.416 Seconds
For Simplify[], I have observed ratios of up to 20:1. All operations
I have timed so far are slower. Perhaps the Mac kernel has not been
properly optimized. Mma 3.0 on a NT HP 200MHz Pentium seems noticeably
faster although the front-end is buggier.
What machine and OS were used in those tests?