You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@mahout.apache.org by Saikat Kanjilal <sx...@hotmail.com> on 2014/09/13 19:14:15 UTC
Affinity matrix computation
Hi Committers,I'm beginning some work on the affinity matrix computation in mahout-dsl, I was wondering where in the directory structure I should put this effort, are we placing all our algorithms in mahout-dsl in a specific area?Thanks in advance.
RE: Affinity matrix computation
Posted by Andrew Palumbo <ap...@outlook.com>.
but again- is up for discussion, and I'm just catching up on todays emails, and pat's Idea of a scala-core module is making sense.. for now math-scala though probably
From: ap.dev@outlook.com
To: dev@mahout.apache.org
Subject: RE: Affinity matrix computation
Date: Sat, 13 Sep 2014 13:41:24 -0400
I think that this is something that is currently up for discussion. But if it is "pure" Dsl code maybe o.a.m.clustering in the math-scala package.
> From: sxk1969@hotmail.com
> To: dev@mahout.apache.org
> Subject: Affinity matrix computation
> Date: Sat, 13 Sep 2014 10:14:15 -0700
>
> Hi Committers,I'm beginning some work on the affinity matrix computation in mahout-dsl, I was wondering where in the directory structure I should put this effort, are we placing all our algorithms in mahout-dsl in a specific area?Thanks in advance.
RE: Affinity matrix computation
Posted by Andrew Palumbo <ap...@outlook.com>.
I think that this is something that is currently up for discussion. But if it is "pure" Dsl code maybe o.a.m.clustering in the math-scala package.
> From: sxk1969@hotmail.com
> To: dev@mahout.apache.org
> Subject: Affinity matrix computation
> Date: Sat, 13 Sep 2014 10:14:15 -0700
>
> Hi Committers,I'm beginning some work on the affinity matrix computation in mahout-dsl, I was wondering where in the directory structure I should put this effort, are we placing all our algorithms in mahout-dsl in a specific area?Thanks in advance.
RE: Affinity matrix computation
Posted by Andrew Palumbo <ap...@outlook.com>.
ahh sorry- didn't realize this..
> Date: Sat, 13 Sep 2014 13:41:11 -0400
> From: squinn@gatech.edu
> To: dev@mahout.apache.org
> Subject: Re: Affinity matrix computation
>
> Since it's an input processing method--rather than strictly an algorithm
> in the category of SVMs, K-means, etc--and since you're early in the
> development cycle, wherever makes it easiest is probably best for now.
> We can always merge it elsewhere once you're ready to submit a PR.
>
> Of course someone please correct me if I'm mistaken.
>
> On 9/13/14, 1:14 PM, Saikat Kanjilal wrote:
> > Hi Committers,I'm beginning some work on the affinity matrix computation in mahout-dsl, I was wondering where in the directory structure I should put this effort, are we placing all our algorithms in mahout-dsl in a specific area?Thanks in advance.
>
Re: Affinity matrix computation
Posted by Shannon Quinn <sq...@gatech.edu>.
Since it's an input processing method--rather than strictly an algorithm
in the category of SVMs, K-means, etc--and since you're early in the
development cycle, wherever makes it easiest is probably best for now.
We can always merge it elsewhere once you're ready to submit a PR.
Of course someone please correct me if I'm mistaken.
On 9/13/14, 1:14 PM, Saikat Kanjilal wrote:
> Hi Committers,I'm beginning some work on the affinity matrix computation in mahout-dsl, I was wondering where in the directory structure I should put this effort, are we placing all our algorithms in mahout-dsl in a specific area?Thanks in advance.