[R] Error with KNN

[Uwe Ligges]

On 23.02.2010 05:34, Nancy Adam wrote:
>
>
>
>
> Hi Uwe,
> Thank you for your email.
> I need to compute one value indicating to the performance of knn but I don’t know anything other than MSE. Can you please tell me how I do that?


Hi Nancy,

since knn is a classification method, I think MSE is a rather untypical 
measure. It is more usual to use a misclassification rate based on cross 
validation, bootstrap - or at least a test data set.

See packages "caret" and "ipred" for functions that make it easy to 
calculate a cross validated misclassification error for knn on your 
data, for example.

Best wishes,
Uwe



> Many thanks,
> Nancy
>
>
>> Date: Mon, 22 Feb 2010 17:19:48 +0100
>> From: ligges at statistik.tu-dortmund.de
>> To: nancyadam84 at hotmail.com
>> CC: r-help at r-project.org
>> Subject: Re: [R] Error with KNN
>>
>>
>>
>> On 20.02.2010 19:10, Nancy Adam wrote:
>>>
>>>
>>>
>>> Hi all
>>>
>>> I receive this error message: “Error in mymodel$MSE : $ operator is invalid for atomic vectors”
>>> When I use KNN.
>>> This is the code:
>>>
>>> library(e1071)
>>> train<- rbind(iris3[,,1], iris3[,,2], iris3[,,3])
>>> cl<- factor(c(rep("s",50), rep("c",50), rep("v",50)))
>>> knn.cv(train, cl, k = 3, prob = TRUE)
>>> sqrt(mymodel$MSE)
>>> #error message : “Error in mymodel$MSE : $ operator is invalid for atomic vectors”
>>>
>>
>> There is no component MSE in mymodel. Hence you cannot extract it.
>>
>> Uwe Ligges
>>
>>
>>
>>
>>
>>> thanks,
>>> Nancy
>>>
>>> _________________________________________________________________
>>> Hotmail: Trusted email with powerful SPAM protection.
>>>
>>> [[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.
>   		 	   		
> _________________________________________________________________
> Hotmail: Free, trusted and rich email service.
> https://signup.live.com/signup.aspx?id=60969