CODEWITHSUNDEEP
rr-caretconfusion-matrix
ForEverNewbie
ForEverNewbie

Reputation: 512

How to extract predictions and actual values vectors from a matrix to use them with confusionMatrix() in R?

Say i have the matrix

> a <- matrix(c(1,2,3,4,5,6,7,8,9),nrow=3)
> rownames(a)=c('A','B','C')
> colnames(a)=c('A','B','C')
> a
  A B C
A 1 4 7
B 2 5 8
C 3 6 9

Considering that columns would represent the actual class and rows the predicted class, how can I extract the predictions and actual values vectors to use them in confusionMatrix()?.

Upvotes: 1

Views: 584

Answers (1)

StupidWolf
StupidWolf

Reputation: 46968

I guess you are referring to confusionMatrix() from caret. This is already a confusion matrix, and you can just pass the predictions into the function using as.table(), see example, where we set up a model and train / test data:

library(caret)
set.seed(111)
idx = sample(1:nrow(iris),100)
trainData = iris[idx,]
testData = iris[-idx,]

mdl = train(Species ~ .,data=trainData,
method="rf",trControl=trainControl(method="cv"))
pred = predict(mdl,testData)
actual = testData$Species

Confusion matrix with labels:

confusionMatrix(pred,actual)
Confusion Matrix and Statistics

            Reference
Prediction   setosa versicolor virginica
  setosa         20          0         0
  versicolor      0         11         2
  virginica       0          0        17

Confusion matrix with table or a matrix :

a = matrix(table(pred,actual),nrow=3)
colnames(a) = levels(testData$Species)
rownames(a) = levels(testData$Species)

           setosa versicolor virginica
setosa         20          0         0
versicolor      0         11         2
virginica       0          0        17

confusionMatrix(as.table(a))

Confusion Matrix and Statistics

           setosa versicolor virginica
setosa         20          0         0
versicolor      0         11         2
virginica       0          0        17

Overall Statistics
                                          
               Accuracy : 0.96            
                 95% CI : (0.8629, 0.9951)
    No Information Rate : 0.4             
    P-Value [Acc > NIR] : < 2.2e-16

If you really need them in a vector, (this sounds super bizarre for me) using a:

actual_vector = rep(colnames(a),colSums(a))
pred_vector = rep(rownames(a),rowSums(a))

table(actual_vector) == table(actual)
actual_vector
    setosa versicolor  virginica 
      TRUE       TRUE       TRUE 

table(pred_vector) == table(pred)
pred_vector
    setosa versicolor  virginica 
      TRUE       TRUE       TRUE

Upvotes: 2

Related Questions