| Re: [eigen] Eigen is slow with complex matrices |
[ Thread Index |
Date Index
| More lists.tuxfamily.org/eigen Archives
]
- To: eigen@xxxxxxxxxxxxxxxxxxx
- Subject: Re: [eigen] Eigen is slow with complex matrices
- From: Benoit Jacob <jacob.benoit.1@xxxxxxxxx>
- Date: Tue, 20 Oct 2009 13:41:16 -0400
- Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:mime-version:received:in-reply-to:references :date:message-id:subject:from:to:content-type :content-transfer-encoding; bh=8En/QJ7A/50YpREbjtlEUsA7Xk75ukmemuo+GCHJltA=; b=u1zIiG0m8SMS0apSaZ4bYSZ+zLVnF27MRJzq3dmJ/aybHaMrBTWTqKZoye4UmYqiOK HE8or56L832IP1GC9GJFkw4aL2kg0PTxbRv8Y7qPWBvikSFWDmGrkxkXJkYSwMlJbYQm MV8LOl4XpAeJvlNfB4BA/N6JUhzb40e+Zc2w4=
- Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :content-type:content-transfer-encoding; b=jM6X5pOis0JE/aWZuw3oBHrH3ur27OVGdRbNphM+VevkQi2KME766HCE3FKNi2JjGL Fb1uCDDc2wfeDo9NXTnQh6M9GGjc0mez2o6cUcC6XzN1lmUglkvjmVuuH1QWdfqZB7/r RK4WcEbhQDOTmapNdvq9j1Ar1eygorKXhJeLc=
2009/10/20 Rohit Garg <rpg.314@xxxxxxxxx>:
> On Tue, Oct 20, 2009 at 9:45 PM, Mark Borgerding <mark@xxxxxxxxxxxxxx> wrote:
>
>> Good thing Rohit has ECC memory.
>
> :)
>
> Any way, I just want to point out that we had a big solve() API debate
> as well, where we decided that passing arguments by pointer is
> undesirable. So please avoid, if possible, pass-by-pointer in the
> public API. Internally of course, we can choose whatever calling
> convention seems appropriate.
Right, two remarks though:
- the results of that debate are only getting implemented now (see the
refactor_solve_api fork)
- the replacement needs to be decided on a per-case basis, sometimes
it's passing by reference,
sometime it's returning a proxy ReturnByValue object...
so to make it short, just send in the sauce and we'll work on it
together. That shouldn't block the inclusion of FFT into our devel
branch.
Benoit
>>
>> FYI: I had to look at the email I sent on June 17,2009. It gave a bit of
>> the background on what
>> http://bitbucket.org/mborgerd/eigen2_for_fft/src/tip/unsupported/Eigen/Complex
>> tries to accomplish.
>>
>>
>>
>>>
>>>>
>>>> What happened to it?
>>>>
>>>
>>> It's still there.
>>>
>>>
>>
>> And it is separable from any of the FFT work. It can be merged even if FFT
>> is not.
>>
>>
>> I'm not familiar with merging in hg. Can anybody send me a link on how I
>> can merge the eigen2_for_fft branch back into the main line?
>> Do I even have permissions to do so?
>>
>> -- Mark
>>
>>
>>
>>
>
>
>
> --
> Rohit Garg
>
> http://rpg-314.blogspot.com/
>
> Senior Undergraduate
> Department of Physics
> Indian Institute of Technology
> Bombay
>
>
>