You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@mahout.apache.org by Keith Thompson <kt...@binghamton.edu> on 2011/05/23 23:22:01 UTC

Random Forest from Command Line

I noticed the classes org.apache.mahout.df.mapreduce.inmem and
org.apache.mahout.df.mapreduce.partial are available for Random Decision
Forests; however, when I run bin/mahout these are not listed as available
command line options.  Does that mean these classification algorithms must
be called from within a hadoop job rather than run from the command line?

Re: Re : Random Forest from Command Line

Posted by Ted Dunning <te...@gmail.com>.
We all start there.

On Tue, May 24, 2011 at 10:26 AM, Keith Thompson <kt...@binghamton.edu>wrote:

> Unfortunately, I'm not only a newbie to Hadoop and
> Mahout but to Linux as well, so I am learning everything all at once.
>

Re: Re : Random Forest from Command Line

Posted by Keith Thompson <kt...@binghamton.edu>.
Thank you.   I realize I originally exported M2_HOME in the command terminal
but didn't add it to the path in my ~/.bashrc file.  After doing that, it
seems to run fine. :)  Unfortunately, I'm not only a newbie to Hadoop and
Mahout but to Linux as well, so I am learning everything all at once.

On Tue, May 24, 2011 at 1:13 PM, Ted Dunning <te...@gmail.com> wrote:

> This kind of strongly indicates that you don't have mvn installed
> correctly.
>
> On Tue, May 24, 2011 at 10:10 AM, Keith Thompson <kthomps6@binghamton.edu
> >wrote:
>
> > What does [which mvn] say?
> > which: no mvn in
> >
> >
> (/usr/lib/mpi/gcc/openmpi/bin:/home/k_thomp/bin:/usr/local/bin:/usr/bin:/bin:/usr/bin/X11:/usr/X11R6/bin:/usr/games:/opt/kde3/bin:/usr/lib/mit/bin:/usr/lib/mit/sbin)
>

Re: Re : Random Forest from Command Line

Posted by Ted Dunning <te...@gmail.com>.
This kind of strongly indicates that you don't have mvn installed correctly.

On Tue, May 24, 2011 at 10:10 AM, Keith Thompson <kt...@binghamton.edu>wrote:

> What does [which mvn] say?
> which: no mvn in
>
> (/usr/lib/mpi/gcc/openmpi/bin:/home/k_thomp/bin:/usr/local/bin:/usr/bin:/bin:/usr/bin/X11:/usr/X11R6/bin:/usr/games:/opt/kde3/bin:/usr/lib/mit/bin:/usr/lib/mit/sbin)

Re: Re : Random Forest from Command Line

Posted by Keith Thompson <kt...@binghamton.edu>.
Thanks.  I do appreciate your time in looking at it.

Which platform again?
I am using OpenSuse 11.2.

What does [which mvn] say?
which: no mvn in
(/usr/lib/mpi/gcc/openmpi/bin:/home/k_thomp/bin:/usr/local/bin:/usr/bin:/bin:/usr/bin/X11:/usr/X11R6/bin:/usr/games:/opt/kde3/bin:/usr/lib/mit/bin:/usr/lib/mit/sbin)

On Tue, May 24, 2011 at 1:06 PM, Ted Dunning <te...@gmail.com> wrote:

> That is confusing.
>
> Which platform again?
>
> What does [which mvn] say?
>
> On Tue, May 24, 2011 at 10:02 AM, Keith Thompson <kthomps6@binghamton.edu
> >wrote:
>
> > But it worked because I used it to install Mahout in the first place ...
> >
> > On Tue, May 24, 2011 at 12:47 PM, Ted Dunning <te...@gmail.com>
> > wrote:
> >
> > > It sounds like maven is not correctly installed on your machine.
> > >
> > > On Tue, May 24, 2011 at 9:45 AM, Keith Thompson <
> kthomps6@binghamton.edu
> > > >wrote:
> > >
> > > > When I go to my MAHOUT_HOME, it says command mvn not found.
> > >
> >
>

Re: Re : Random Forest from Command Line

Posted by Ted Dunning <te...@gmail.com>.
That is confusing.

Which platform again?

What does [which mvn] say?

On Tue, May 24, 2011 at 10:02 AM, Keith Thompson <kt...@binghamton.edu>wrote:

> But it worked because I used it to install Mahout in the first place ...
>
> On Tue, May 24, 2011 at 12:47 PM, Ted Dunning <te...@gmail.com>
> wrote:
>
> > It sounds like maven is not correctly installed on your machine.
> >
> > On Tue, May 24, 2011 at 9:45 AM, Keith Thompson <kthomps6@binghamton.edu
> > >wrote:
> >
> > > When I go to my MAHOUT_HOME, it says command mvn not found.
> >
>

Re: Re : Random Forest from Command Line

Posted by Keith Thompson <kt...@binghamton.edu>.
But it worked because I used it to install Mahout in the first place ...

On Tue, May 24, 2011 at 12:47 PM, Ted Dunning <te...@gmail.com> wrote:

> It sounds like maven is not correctly installed on your machine.
>
> On Tue, May 24, 2011 at 9:45 AM, Keith Thompson <kthomps6@binghamton.edu
> >wrote:
>
> > When I go to my MAHOUT_HOME, it says command mvn not found.
>

Re: Re : Random Forest from Command Line

Posted by Ted Dunning <te...@gmail.com>.
It sounds like maven is not correctly installed on your machine.

On Tue, May 24, 2011 at 9:45 AM, Keith Thompson <kt...@binghamton.edu>wrote:

> When I go to my MAHOUT_HOME, it says command mvn not found.

Re: Re : Random Forest from Command Line

Posted by Keith Thompson <kt...@binghamton.edu>.
Hi, Thanks Deneche.  I don't know why that information isn't at
https://cwiki.apache.org/MAHOUT/random-forests.html ... it's a big help.  I
am working through that tutorial now but am stuck at mvn clean install
-Dskiptests.  When I go to my MAHOUT_HOME, it says command mvn not found.
Maybe it's a problem with my file structure.  I followed the instructions
for MAHOUT installation the best I could but maybe I installed it in the
wrong place.  Currently I have this structure:

~/hadoop-0.20.0/bin/hadoop
~/apache-maven-2.2.1/bin/mvn
~/trunk/bin/mahout

When the directions at
https://cwiki.apache.org/MAHOUT/buildingmahout.htmlsaid to change to
the "checked out" directory I assumed that was trunk.  Was
I wrong?

With the installation setup I have, can I get everything to run correctly
still?

On Mon, May 23, 2011 at 11:24 PM, deneche abdelhakim <a_...@yahoo.fr>wrote:

> you can use Mahout's Random Forests from the command line, take a look at
> the following :
>
> https://cwiki.apache.org/MAHOUT/partial-implementation.html
>
>
>
> ________________________________
> De : Keith Thompson <kt...@binghamton.edu>
> À : user@mahout.apache.org
> Envoyé le : Lundi 23 Mai 2011 22h22
> Objet : Random Forest from Command Line
>
> I noticed the classes org.apache.mahout.df.mapreduce.inmem and
> org.apache.mahout.df.mapreduce.partial are available for Random Decision
> Forests; however, when I run bin/mahout these are not listed as available
> command line options.  Does that mean these classification algorithms must
> be called from within a hadoop job rather than run from the command line?
>

Re : Random Forest from Command Line

Posted by deneche abdelhakim <a_...@yahoo.fr>.
you can use Mahout's Random Forests from the command line, take a look at the following :

https://cwiki.apache.org/MAHOUT/partial-implementation.html



________________________________
De : Keith Thompson <kt...@binghamton.edu>
À : user@mahout.apache.org
Envoyé le : Lundi 23 Mai 2011 22h22
Objet : Random Forest from Command Line

I noticed the classes org.apache.mahout.df.mapreduce.inmem and
org.apache.mahout.df.mapreduce.partial are available for Random Decision
Forests; however, when I run bin/mahout these are not listed as available
command line options.  Does that mean these classification algorithms must
be called from within a hadoop job rather than run from the command line?

Re: Random Forest from Command Line

Posted by Ted Dunning <te...@gmail.com>.
I think it just means that those were reference implementations intended to
be a benchmark for the parallel implementation.

On Mon, May 23, 2011 at 2:22 PM, Keith Thompson <kt...@binghamton.edu>wrote:

> Does that mean these classification algorithms must
> be called from within a hadoop job rather than run from the command line?
>