e1071 SVM: Error trying to predict - r

I keep receiving this error and I cannot figure out why.
Error in scale.default(newdata[, object$scaled, drop = FALSE], center
= object$x.scale$"scaled:center", : length of 'center' must equal the number of columns of 'x'
I'm using the default iris dataset, and here is all of my code. It's an attempt at implementing a multiclass SVM using the pairwise method.
# pass in the dataframe & the number of classes
multiclass.svm <- function(data) {
class.vec = data[,length(data)]
levels = levels(class.vec)
pair1 <- data[which(class.vec == levels[1]),]
pair1 <- droplevels(pair1)
pair2 <- data[which(class.vec == levels[length(levels)]),]
pair2 <- droplevels(pair2)
pairs = list(rbind(pair1, pair2))
# print(pairs)
for(i in 2:length(levels)){
L1 <- data[which(class.vec == levels[i-1]),]
L1 <- droplevels(L1)
L2 <- data[which(class.vec == levels[i]),]
L2 <- droplevels(L2)
pair <- list(rbind(L1, L2))
pairs <- c(pairs, pair)
}
# now we construct our (n choose 2) binary models
models = list()
for(pair in pairs){
classifier = pair[,length(pair)]
p.svm = svm(formula=classifier~., data=pair)
models = c(models, list(p.svm))
}
for(model in models){
test = iris[1,]
print(predict(model, test))
}
return(models)
}
Testing/usage:
> h = multiclass.svm(iris)
Show Traceback
Rerun with Debug
Error in scale.default(newdata[, object$scaled, drop = FALSE], center = object$x.scale$"scaled:center", :
length of 'center' must equal the number of columns of 'x'
>
Any help would be very much appreciated... I've found a few other questions on this very topic to no avail. Thank you.

Okay, so the answer is unfortunately quite tricky (depending on your dataset.) The problem is that in the iris dataset, there are THREE levels of classification. Since I'm breaking the classes into pairs, each of my models only have TWO levels of classification.
When using predict the model you've trained on AND the value you're testing must both have the same levels. So, the tricky part (at least in this case) is deleting the unnecessary levels from each pair.
I recommend using the library plyr for its revalue function. To remove specific levels (instead of all unused a la the drop levels function) you can use revalue and rename each unwanted level to an existing one (essentially destroying it.)
Credit to this polish blogger for steering me in the right direction:
http://ppiotrow.blogspot.com/2013/04/solved-r-svm-test-data-does-not-match.html
The quick and easy way, though, to solve my specific problem was simply removing all of the droplevels calls haha. Since the SVM won't find any points to draw upon the unused level, there is no actual problem with leaving the extraneous level in.
Hope this helps someone out there.
Mike

I have experienced the same issue. I fixed the error by converting all of the predictors in the test set to their correct class, i.e., as.factor, as.numeric.
For example, if a numeric predictor variable in the training set is in memory as a character variable in your test set you will get this error. I hope this helps.

Related

Error with RandomForest in R because of "too many categories"

I'm trying to train a RF model in R, but when i try to define the model:
rf <- randomForest(labs ~ .,data=as.matrix(dd.train))
It gives me the error:
Error in randomForest.default(m, y, ...) :
Can not handle categorical predictors with more than 53 categories.
Any idea what could it be?
And no, before you say "You have some categoric variable with more than 53 categories". No, all variables but labs are numeric.
Tim Biegeleisen: Read the last line of my question and you will see why is not the same as the one you are linking!
Edited to address followup from OP
I believe using as.matrix in this case implicitly creates factors. It is also not necessary for this packages. You can keep it as a data frame, but will need to make sure that any unused factor levels are dropped by using droplevels (or something similar). There are many reasons an unused factor may be in your data set, but a common one is a dropped observation.
Below is a quick example that reproduces your error:
library('randomForest')
#making a toy data frame
x <- data.frame('one' = c(1,1,1,1,1,seq(50) ),
'two' = c(seq(54),NA),
'three' = seq(55),
'four' = seq(55) )
x$one <- as.factor(x$one)
x <- na.omit(x) #getting rid of an NA. Note this removes the whole row.
randomForest(one ~., data = as.matrix(x)) #your first error
randomForest(one ~., data = x) #your second error
x <- droplevels(x)
randomForest(one ~., data = x) #OK

Kaggle Digit Recognizer Using SVM (e1071): Error in predict.svm(ret, xhold, decision.values = TRUE) : Model is empty

I am trying to solve the digit Recognizer competition in Kaggle and I run in to this error.
I loaded the training data and adjusted the values of it by dividing it with the maximum pixel value which is 255. After that, I am trying to build my model.
Here Goes my code,
Given_Training_data <- get(load("Given_Training_data.RData"))
Given_Testing_data <- get(load("Given_Testing_data.RData"))
Maximum_Pixel_value = max(Given_Training_data)
Tot_Col_Train_data = ncol(Given_Training_data)
training_data_adjusted <- Given_Training_data[, 2:ncol(Given_Training_data)]/Maximum_Pixel_value
testing_data_adjusted <- Given_Testing_data[, 2:ncol(Given_Testing_data)]/Maximum_Pixel_value
label_training_data <- Given_Training_data$label
final_training_data <- cbind(label_training_data, training_data_adjusted)
smp_size <- floor(0.75 * nrow(final_training_data))
set.seed(100)
training_ind <- sample(seq_len(nrow(final_training_data)), size = smp_size)
training_data1 <- final_training_data[training_ind, ]
train_no_label1 <- as.data.frame(training_data1[,-1])
train_label1 <-as.data.frame(training_data1[,1])
svm_model1 <- svm(train_label1,train_no_label1) #This line is throwing an error
Error : Error in predict.svm(ret, xhold, decision.values = TRUE) : Model is empty!
Please Kindly share your thoughts. I am not looking for an answer but rather some idea that guides me in the right direction as I am in a learning phase.
Thanks.
Update to the question :
trainlabel1 <- train_label1[sapply(train_label1, function(x) !is.factor(x) | length(unique(x))>1 )]
trainnolabel1 <- train_no_label1[sapply(train_no_label1, function(x) !is.factor(x) | length(unique(x))>1 )]
svm_model2 <- svm(trainlabel1,trainnolabel1,scale = F)
It didn't help either.
Read the manual (https://cran.r-project.org/web/packages/e1071/e1071.pdf):
svm(x, y = NULL, scale = TRUE, type = NULL, ...)
...
Arguments:
...
x a data matrix, a vector, or a sparse matrix (object of class
Matrix provided by the Matrix package, or of class matrix.csr
provided by the SparseM package,
or of class simple_triplet_matrix provided by the slam package).
y a response vector with one label for each row/component of x.
Can be either a factor (for classification tasks) or a numeric vector
(for regression).
Therefore, the mains problems are that your call to svm is switching the data matrix and the response vector, and that you are passing the response vector as integer, resulting in a regression model. Furthermore, you are also passing the response vector as a single-column data-frame, which is not exactly how you are supposed to do it. Hence, if you change the call to:
svm_model1 <- svm(train_no_label1, as.factor(train_label1[, 1]))
it will work as expected. Note that training will take some minutes to run.
You may also want to remove features that are constant (where the values in the respective column of the training data matrix are all identical) in the training data, since these will not influence the classification.
I don't think you need to scale it manually since svm itself will do it unlike most neural network package.
You can also use the formula version of svm instead of the matrix and vectors which is
svm(result~.,data = your_training_set)
in your case, I guess you want to make sure the result to be used as factor,because you want a label like 1,2,3 not 1.5467 which is a regression
I can debug it if you can share the data:Given_Training_data.RData

Subscript out of bound error in predict function of randomforest

I am using random forest for prediction and in the predict(fit, test_feature) line, I get the following error. Can someone help me to overcome this. I did the same steps with another dataset and had no error. but I get error here.
Error: Error in x[, vname, drop = FALSE] : subscript out of bounds
training_index <- createDataPartition(shufflled[,487], p = 0.8, times = 1)
training_index <- unlist(training_index)
train_set <- shufflled[training_index,]
test_set <- shufflled[-training_index,]
accuracies<- c()
k=10
n= floor(nrow(train_set)/k)
for(i in 1:k){
sub1<- ((i-1)*n+1)
sub2<- (i*n)
subset<- sub1:sub2
train<- train_set[-subset, ]
test<- train_set[subset, ]
test_feature<- test[ ,-487]
True_Label<- as.factor(test[ ,487])
fit<- randomForest(x= train[ ,-487], y= as.factor(train[ ,487]))
prediction<- predict(fit, test_feature) #The error line
correctlabel<- prediction == True_Label
t<- table(prediction, True_Label)
}
I had similar problem few weeks ago.
To go around the problem, you can do this:
df$label <- factor(df$label)
Instead of as.factor try just factor generic function. Also, try first naming your label variable.
Are there identical column names in your training and validation x?
I had the same error message and solved it by renaming my column names because my data was a matrix and their colnames were all empty, i.e. "".
Your question is not very clear, anyway I try to help you.
First of all check your data to see the distribution in levels of your various predictors and outcomes.
You may find that some of your predictor levels or outcome levels are very highly skewed, or some outcomes or predictor levels are very rare. I got that error when I was trying to predict a very rare outcome with a heavily tuned random forest, and so some of the predictor levels were not actually in the training data. Thus a factor level appears in the test data that the training data thinks is out of bounds.
Alternatively, check the names of your variables.
Before calling predict() to make sure that the variable names match.
Without your data files, it's hard to tell why your first example worked.
For example You can try:
names(test) <- names(train)
Add the expression
dimnames(test_feature) <- NULL
before
prediction <- predict(fit, test_feature)

predict in caret ConfusionMatrix is removing rows

I'm fairly new to using the caret library and it's causing me some problems. Any
help/advice would be appreciated. My situations are as follows:
I'm trying to run a general linear model on some data and, when I run it
through the confusionMatrix, I get 'the data and reference factors must have
the same number of levels'. I know what this error means (I've run into it before), but I've double and triple checked my data manipulation and it all looks correct (I'm using the right variables in the right places), so I'm not sure why the two values in the confusionMatrix are disagreeing. I've run almost the exact same code for a different variable and it works fine.
I went through every variable and everything was balanced until I got to the
confusionMatrix predict. I discovered this by doing the following:
a <- table(testing2$hold1yes0no)
a[1]+a[2]
1543
b <- table(predict(modelFit,trainTR2))
dim(b)
[1] 1538
Those two values shouldn't disagree. Where are the missing 5 rows?
My code is below:
set.seed(2382)
inTrain2 <- createDataPartition(y=HOLD$hold1yes0no, p = 0.6, list = FALSE)
training2 <- HOLD[inTrain2,]
testing2 <- HOLD[-inTrain2,]
preProc2 <- preProcess(training2[-c(1,2,3,4,5,6,7,8,9)], method="BoxCox")
trainPC2 <- predict(preProc2, training2[-c(1,2,3,4,5,6,7,8,9)])
trainTR2 <- predict(preProc2, testing2[-c(1,2,3,4,5,6,7,8,9)])
modelFit <- train(training2$hold1yes0no ~ ., method ="glm", data = trainPC2)
confusionMatrix(testing2$hold1yes0no, predict(modelFit,trainTR2))
I'm not sure as I don't know your data structure, but I wonder if this is due to the way you set up your modelFit, using the formula method. In this case, you are specifying y = training2$hold1yes0no and x = everything else. Perhaps you should try:
modelFit <- train(trainPC2, training2$hold1yes0no, method="glm")
Which specifies y = training2$hold1yes0no and x = trainPC2.

Use of randomforest() for classification in R?

I originally had a data frame composed of 12 columns in N rows. The last column is my class (0 or 1). I had to convert my entire data frame to numeric with
training <- sapply(training.temp,as.numeric)
But then I thought I needed the class column to be a factor column to use the randomforest() tool as a classifier, so I did
training[,"Class"] <- factor(training[,ncol(training)])
I proceed to creating the tree with
training_rf <- randomForest(Class ~., data = trainData, importance = TRUE, do.trace = 100)
But I'm getting two errors:
1: In Ops.factor(training[, "Status"], factor(training[, ncol(training)])) :
<= this is not relevant for factors (roughly translated)
2: In randomForest.default(m, y, ...) :
The response has five or fewer unique values. Are you sure you want to do regression?
I would appreciate it if someone could point out the formatting mistake I'm making.
Thanks!
So the issue is actually quite simple. It turns out my training data was an atomic vector. So it first had to be converted as a data frame. So I needed to add the following line:
training <- as.data.frame(training)
Problem solved!
First, your coercion to a factor is not working because of syntax errors. Second, you should always use indexing when specifying a RF model. Here are changes in your code that should make it work.
training <- sapply(training.temp,as.numeric)
training[,"Class"] <- as.factor(training[,"Class"])
training_rf <- randomForest(x=training[,1:(ncol(training)-1)], y=training[,"Class"],
importance=TRUE, do.trace=100)
# You can also coerce to a factor directly in the model statement
training_rf <- randomForest(x=training[,1:(ncol(training)-1)], y=as.factor(training[,"Class"]),
importance=TRUE, do.trace=100)

Resources