Path: news.mathworks.com!not-for-mail
From: Ting <Ting_noreply@mathworks.com>
Newsgroups: comp.soft-sys.matlab
Subject: Re: randfeatures
Date: Tue, 14 Oct 2008 14:06:47 -0400
Organization: The MathWorks, Inc.
Lines: 25
Message-ID: <gd2n3n$mg8$1@fred.mathworks.com>
References: <gd2eid$lre$1@fred.mathworks.com>
Reply-To: ting_noreply@mathworks.com
NNTP-Posting-Host: sut.dhcp.mathworks.com
Mime-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-Trace: fred.mathworks.com 1224007607 23048 172.31.57.42 (14 Oct 2008 18:06:47 GMT)
X-Complaints-To: news@mathworks.com
NNTP-Posting-Date: Tue, 14 Oct 2008 18:06:47 +0000 (UTC)
User-Agent: Thunderbird 2.0.0.17 (Windows/20080914)
In-Reply-To: <gd2eid$lre$1@fred.mathworks.com>
Xref: news.mathworks.com comp.soft-sys.matlab:495144


Sunny wrote:
> Hi, 
> 
> Im trying to use the randfeatures fuction, in the example provided by MATLAB it uses two datasets X(1375x60) and GROUP(1x60) 
> 
> example: I = randfeatures(X,GROUP,'SubsetSize',15,'Classifier','da');
> 
> my dataset of the same type (double) is X(1668x1497) and GROUP(1x1497), but, I get the error message "Error using ==> classify
> TRAINING must have more observations than the number of groups."
> 
> I cant work out where this issue lies, my training is 1668 samples which is more than group(60).
> 
> Any help?
> 
> kind regards

Sunny,
Without looking at your data, I don't know what really happened on your 
data. However, the following can be one possible reason.
randfeatures is the bioinformatics toolbox, so every column of X 
represents an observed sample. This is a convention used for 
bioinformatics applications and is clearly documented.

-Ting
Mathworks