Neuralnet Error

Hi -- this isn't really reproducible: can you follow the advice given
here and make a small reproducible example:

http://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example

Best,
Michael

On Wed, Aug 1, 2012 at 12:51 PM, Rahul Bhalla

I require some help in debugging this code

library(neuralnet)
ir<-read.table(file="iris_data.txt",header=TRUE,row.names=NULL)
ir1 <- data.frame(ir[1:100,2:6])
ir2 <- data.frame(ifelse(ir1$Species=="setosa",1,ifelse(ir1$Species=="versicolor",0,"")))
colnames(ir2)<-("Output")
ir3 <- data.frame(rbind(ir1[1:4],ir2))
#rownames(ir3)<-c("SL","SW","PL","PW","Output")
print(ir3)
n<- neuralnet(Output~Sepal.Length+Sepal.Width+Petal.Length+Petal.Width,data=ir3,err.fct="sse",hidden=2,linear.output=FALSE)

Output:

Error in neurons[[i]] %*% weights[[i]] :
  requires numeric/complex matrix/vector arguments

Any assisstance is appreciated
        [[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.

Hello,

I guess the error might be due to the struture of your dataset.? 

Try this:

data(iris)
iris1<-iris
iris1[iris1$Species=="setosa","Output"]<- 1
iris1[iris1$Species=="versicolor","Output"]<- 0
#Here, if you set "virginica" to 2, still it runs, with a warning message "-- response is not binary".? I also run it without setting values for viriginica ,i.e. NA, it didn't gave any error or warnings.? 

iris1[iris1$Species=="virginica","Output"]<- 0
nn1<-neuralnet(Output~Sepal.Length+Sepal.Width+Petal.Length+Petal.Width,data=iris1,hidden=2,err.fct="ce",linear.output=FALSE)
?plot(nn1)
?nn1$result.matrix
?????????????????????????????????????? 1
error???????????????????? 0.018911271401
reached.threshold???????? 0.009459417063
steps??????????????????? 92.000000000000
Intercept.to.1layhid1??? -4.353018230633
Sepal.Length.to.1layhid1? 1.720240396708
Sepal.Width.to.1layhid1? -6.428901709134
Petal.Length.to.1layhid1? 3.273018763978
Petal.Width.to.1layhid1?? 7.110403675648
Intercept.to.1layhid2???? 4.815489181151
Sepal.Length.to.1layhid2 -0.123278321659
Sepal.Width.to.1layhid2?? 6.348589101327
Petal.Length.to.1layhid2 -6.376680199897
Petal.Width.to.1layhid2? -8.065600706230
Intercept.to.Output????? -0.224816723130
1layhid.1.to.Output????? -9.042646714781
1layhid.2.to.Output?????? 8.799061406981


?str(iris1)
'data.frame':??? 150 obs. of? 6 variables:
?$ Sepal.Length: num? 5.1 4.9 4.7 4.6 5 5.4 4.6 5 4.4 4.9 ...
?$ Sepal.Width : num? 3.5 3 3.2 3.1 3.6 3.9 3.4 3.4 2.9 3.1 ...
?$ Petal.Length: num? 1.4 1.4 1.3 1.5 1.4 1.7 1.4 1.5 1.4 1.5 ...
?$ Petal.Width : num? 0.2 0.2 0.2 0.2 0.2 0.4 0.3 0.2 0.2 0.1 ...
?$ Species???? : Factor w/ 3 levels "setosa","versicolor",..: 1 1 1 1 1 1 1 1 1 1 ...
?$ Output????? : num? 1 1 1 1 1 1 1 1 1 1 ...


Hope it helps you.



A.K.


----- Original Message -----
From: Rahul Bhalla <rahulbhalla_3521 at yahoo.com>
To: "r-help at r-project.org" <r-help at r-project.org>
Cc: 
Sent: Wednesday, August 1, 2012 1:51 PM
Subject: [R] Neuralnet Error

I require some help in debugging this code?

library(neuralnet)
ir<-read.table(file="iris_data.txt",header=TRUE,row.names=NULL)
ir1 <- data.frame(ir[1:100,2:6])
ir2 <- data.frame(ifelse(ir1$Species=="setosa",1,ifelse(ir1$Species=="versicolor",0,"")))
colnames(ir2)<-("Output")
ir3 <- data.frame(rbind(ir1[1:4],ir2))
#rownames(ir3)<-c("SL","SW","PL","PW","Output")
print(ir3)
n<- neuralnet(Output~Sepal.Length+Sepal.Width+Petal.Length+Petal.Width,data=ir3,err.fct="sse",hidden=2,linear.output=FALSE)

Output:

Error in neurons[[i]] %*% weights[[i]] :?
? requires numeric/complex matrix/vector arguments

Any assisstance is appreciated
??? [[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.