CODEWITHSUNDEEP
rwekasvm
aceminer
aceminer

Reputation: 4295

Using LibSVM in R

I have a data frame loaded into R.

I am still a beginner with R.

I have installed the e1071 package to use the svm function.

However, i am unsure of which parameters are mapping to which. I would like to have a model with the parameters as per in weka. enter image description here

I do not understand the e1071 svm example as the example does not work. 

 svm(x, y = NULL, scale = TRUE, type = NULL, kernel =
+ "radial", degree = 3, gamma = if (is.vector(x)) 1 else 1 / ncol(x),
+ coef0 = 0, cost = 1, nu = 0.5,
+ class.weights = NULL, cachesize = 40, tolerance = 0.001, epsilon = 0.1,
+ shrinking = TRUE, cross = 0, probability = FALSE, fitted = TRUE,
+ ..., subset, na.action = na.omit)
Error: '...' used in an incorrect context

i currently have this code

x <- read.arff("contact-lenses.arff")

where i loaded my arff into a data frame x.

However, i am unsure of how to continue to use the same parameters as in weka to build the model as in R.

Any help rendered is greatly appreciated.

This is what i tried. 

model <- svm(x, y = NULL, scale = TRUE, type = "C-classification", kernel = "linear")
Error in if (any(co)) { : missing value where TRUE/FALSE needed
In addition: Warning messages:
1: In FUN(newX[, i], ...) : NAs introduced by coercion
2: In FUN(newX[, i], ...) : NAs introduced by coercion
3: In FUN(newX[, i], ...) : NAs introduced by coercion
4: In FUN(newX[, i], ...) : NAs introduced by coercion
5: In FUN(newX[, i], ...) : NAs introduced by coercion

To be honest, i am unsure of some of the parameters like what does it mean by y = NULL. I am literally hitting at random air trying to get it to work.

I have resolved it using the below answer 

model <- svm(`contact-lenses` ~ . , data = x, type = "C-classification", kernel = "linear")

Upvotes: 0

Views: 2499

Answers (1)

bskaggs
bskaggs

Reputation: 1484

The error you get is because the y = NULL means that you aren't actually providing any labels to learn.

Try the following, assuming this is the same contact-lenses.arff file I found via google:

model <- svm(`contact-lenses` ~ . , data = x, type = "C-classification", kernel = "linear")

The first parameter, "`contact-lenses` ~ .", is a formula, and means that you will be predicting the contact-lenses status versus based on all the other features. The backticks around the feature name are necessary because the feature inconveniently has a hyphen in it, which would otherwise be interpreted as subtraction. The data = x parameter means that you will be using x as the underlying data frame to resolve the feature names against.

Upvotes: 2

Related Questions