I am using this function to calculate MCC: https://www.rdocumentation.org/packages/mltools/versions/0.3.5/topics/mcc.

It generates this error:

Error in mcc (actuals = data.test $ Y, preds = pred_rf, TP = tab_rf [1, 1],:
   Either {'preds' and 'actuals'} or {'TP', 'FP', 'TN', 'FN'} or {'confusionM'} should be provided.

My code is the following (I am using Random Forest):

library("mltools")
library("caret") 
library("MASS")
library("MLmetrics") 

data=data.frame()

for (i in 0:4){ 
  
  mu1=c(0 i,4) 
  Sigma= matrix(c(1,0.5,0.5,1),ncol=2)
  X1_distribution=mvrnorm(n=100,mu=mu1,Sigma=Sigma)
  
  
  mu2=c(5-i,1) 
  Sigma= matrix(c(1,0.5,0.5,1),ncol=2) 
  X2_distribution=mvrnorm(n=100,mu=mu2,Sigma=Sigma)
  
  X=rbind(X1_distribution,X2_distribution)
  Y=c(rep(1,100),
      rep(0,100)) 
  data.aux=data.frame(Y=Y,
                      X1=X[,1],
                      X2=X[,2])
  
  data=rbind(data,data.aux)
}

data.train=data[1:800,]
data.test=data[801:1000,]

method_rf= train(form= formula('factor(Y) ~ .'),data=data.train, method="rf") #randomForest
pred_rf=predict(method_rf,data.test[,2:3]) #Predict randomForest
tab_rf=table(actual=data.test$Y,  #ConfusionMatrix for RF
             predicted=pred_rf)

metrics_rf=data.frame(MCC_rf=mcc(actuals=data.test$Y,
                                 preds=pred_rf, 
                                 TP=tab_rf[1,1],
                                 FP=tab_rf[1,2],
                                 TN=tab_rf[2,2],
                                 FN=tab_rf[2,1],
                                 confusionM=tab_rf))
metrics_rf

CodePudding user response：

As the error message implies, you should use either {'preds' and 'actuals'} or {'TP', 'FP', 'TN', 'FN'} or {'confusionM'}, but certainly not all of them togetner, as you are trying to do here.

Changing the MCC calculation in your code to

MCC_rf=mcc(TP=tab_rf[1,1],
           FP=tab_rf[1,2],
           TN=tab_rf[2,2],
           FN=tab_rf[2,1])

works like a charm:

> MCC_rf
[1] 0.60012