Message-ID: <529993C8.6020608@stats.ox.ac.uk>
Date: 2013-11-30T07:29:12Z
From: Brian Ripley
Subject: bnlearn and very large datasets (> 1 million observations)
In-Reply-To: <CADnN+ckp+afLvdOkcsH_X4TNd1erUcBL0KzB5o5JHJn58LPsRQ@mail.gmail.com>
On 30/11/2013 04:52, Jejo Koola wrote:
> Hi
>
> Anyone have experience with very large datasets and the Bayesian Network
> package, bnlearn? In my experience R doesn't react well to very large
> datasets.
Maybe, but a million is not 'very large': R handles billions of
observations without problems on machines with commensurate resources.
Package bnlearn is not 'R'. Your questions are not about R itself and
should be addressed to the package maintainer.
> Is there a way to divide up the dataset into pieces and incrementally learn
> the network with the pieces? This would also be helpful incase R crashes,
> because I could save the network after learning each piece.
>
> Thank you.
>
> [[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>
PLEASE do, including what it says about HTML mail and about 'crashes'.
--
Brian D. Ripley, ripley at stats.ox.ac.uk
Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/
University of Oxford, Tel: +44 1865 272861 (self)
1 South Parks Road, +44 1865 272866 (PA)
Oxford OX1 3TG, UK Fax: +44 1865 272595