The data below is just an example, it is operations on this , or any, data which I am confused about:
library(caret)
set.seed(3433)
data(AlzheimerDisease)
complete <- data.frame(diagnosis, predictors)
in_train <- createDataPartition(complete$diagnosis, p = 0.75)[[1]]
training <- complete[in_train,]
testing <- complete[-in_train,]
predIL <- grep("^IL", names(training))
smalltrain <- training[, c(1, predIL)]
fit_noPCA <- train(diagnosis ~ ., method = "glm", data = smalltrain)
pre_proc_obj <- preProcess(smalltrain[,-1], method = "pca", thresh = 0.8)
smalltrainsPCs <- predict(pre_proc_obj, smalltrain[,-1])
fit_PCA <- train(x = smalltrainsPCs, y = smalltrain$diagnosis, method = "glm")
fit_noPCA$results$Accuracy
fit_PCA$results$Accuracy
When running this code, I get a 0.689539 accuracy for fit_noPCA and 0.682951 accuracy for fit_PCA. But when I rerun the last portion of the code:
fit_noPCA <- train(diagnosis ~ ., method = "glm", data = smalltrain)
pre_proc_obj <- preProcess(smalltrain[,-1], method = "pca", thresh = 0.8)
smalltrainsPCs <- predict(pre_proc_obj, smalltrain[,-1])
fit_PCA <- train(x = smalltrainsPCs, y = smalltrain$diagnosis, method = "glm")
fit_noPCA$results$Accuracy
fit_PCA$results$Accuracy
Then each time I rerun these 6 lines I get different accuracy values. Why is this so? Is it because I am not resetting the seed? Even if, where is the inherent randomness of this process?
By default, the model is trained using bootstrap, you can see it here:
library(caret)
library(AppliedPredictiveModeling)
> fit_noPCA
Generalized Linear Model
251 samples
12 predictor
2 classes: 'Impaired', 'Control'
No pre-processing
Resampling: Bootstrapped (25 reps)
Summary of sample sizes: 251, 251, 251, 251, 251, 251, ...
Resampling results:
Accuracy Kappa
0.6870006 0.04107016
So with every train , the bootstrapped samples will be different, to get back the same result, you can set the seed before running train:
set.seed(111)
fit_PCA <- train(x = smalltrainsPCs, y = smalltrain$diagnosis, method = "glm",trControl=trainControl(method="boot",number=100))
fit_PCA$results$Accuracy
[1] 0.6983512
set.seed(112)
fit_PCA <- train(x = smalltrainsPCs, y = smalltrain$diagnosis, method = "glm",trControl=trainControl(method="boot",number=100))
fit_PCA$results$Accuracy
[1] 0.6991537
set.seed(111)
fit_PCA <- train(x = smalltrainsPCs, y = smalltrain$diagnosis, method = "glm",trControl=trainControl(method="boot",number=100))
fit_PCA$results$Accuracy
[1] 0.6983512
Or use for example cv where you can define the folds using index= in trainControl