The current occupancy dataset as described in Chapter 1, Getting Started, is used to demonstrate the autoencoder setup in R using TensorFlow:

1. Set up the R TensorFlow environment.
2. The load_occupancy_data function can be used to load the data by setting the correct working directory path using setwd:

# Function to load Occupancy data
load_occupancy_data<-function(train){
xFeatures = c("Temperature", "Humidity", "Light", "CO2",
  "HumidityRatio")
yFeatures = "Occupancy"
  if(train){
    occupancy_ds <-  as.matrix(read.csv("datatraining.txt",stringsAsFactors = T))
  } else
  {
    occupancy_ds <- as.matrix(read.csv("datatest.txt",stringsAsFactors = T))
  }
  occupancy_ds<-apply(occupancy_ds[, c(xFeatures, yFeatures)], 2, FUN=as.numeric) 
  return(occupancy_ds)
}

3. The train and test occupancy dataset can be loaded to the R environment with the following script:

occupancy_train <-load_occupancy_data(train=T)
occupancy_test <- load_occupancy_data(train = F)