R/stftEOL.R
In jcpetkovich/rspatiotemp:
library(h2o)
library(doMC)
library(foreach)
h2o.init()
#' Compute the stft of the inputted data and smooth it using loess regression
#' @title STFT loess regression (get.fftloess)
#' @param data The time series data to be transformed
#' @param groupSize The size of the groups that will be fft'd
#' @param scale TRUE or FALSE whether or not to scale the output
#' @return The smoothed stft of the inputted data
#' @export
get.fftloess <- function(data, groupSize,scale){
boundSeq = seq(from=1,to = (length(data)+1), by = groupSize)
len = length(boundSeq) - 1
time = 1:groupSize
# lowerBound = boundSeq[1]
# upperBound = boundSeq[2] - 1
# fft = abs(fft(data[lowerBound:upperBound]))
# lo = predict(loess(fft~time))
registerDoMC()
lo = foreach(i = 1:len,.combine = 'rbind',.inorder = TRUE) %dopar% {
lowerBound = boundSeq[i]
upperBound = boundSeq[i+1] - 1
fft = abs(fft(data[lowerBound:upperBound]))
predict(loess(fft~time))
}
# for(i in 2:len){
# lowerBound = boundSeq[i]
# upperBound = boundSeq[i+1] - 1
# fft = abs(fft(data[lowerBound:upperBound]))
# lo1 = predict(loess(fft~time))
# lo = rbind(lo,lo1)
# }
if(scale)
return(scale(unname(lo)))
return(unname(lo))
}
#' Compute the stft of the inputted data, smooth it using loess regression, but only take a set number of stft pieces
#' @title pieces of STFT loess regression (get.fftloess)
#' @param data The time series data to be transformed
#' @param groupSize The size of the groups that will be fft'd
#' @param smoothSampleNum The number of stft pieces to take
#' @return The smoothed stft of the inputted data
#' @export
get.fftloess.skip <- function(data, groupSize,smoothSampleNum){
boundSeq = seq(from=1,to = (length(data)+1), by = groupSize)
len = length(boundSeq) - 1
boundSeq2 = seq(from = 1, to = length(data),length.out = smoothSampleNum)
time = 1:groupSize
lowerBound = boundSeq[1]
upperBound = boundSeq[2] - 1
fft = abs(fft(data[lowerBound:upperBound]))
lo = predict(loess(fft~time))
for(i in 2:len){
lowerBound = boundSeq[i]
upperBound = boundSeq[i+1] - 1
fft = abs(fft(data[lowerBound:upperBound]))
lo1 = predict(loess(fft~time))[boundSeq2]
lo = rbind(lo,lo1)
}
return(lo)
}
#' Create a deep learning model of smoothed stft data with degradation starting at a chosen percentage to predict a logical variable
#' @title Create Logical Deep Learning Model (createModel.h2o.logic)
#' @param data The time series data to be transformed and used to train the model
#' @param groupSize The size of the groups that will be fft'd
#' @param degradStartPercent The percentage of the data to pass until the system begins to degrade
#' @return A deep learning model of the data to predict a logical output
#' @export
createModel.h2o.logic <- function(data,groupSize,degradStartPercent,scale){
fftloess = get.fftloess(data,groupSize,scale)
rul = c(rep(TRUE,as.integer(nrow(fftloess)*degradStartPercent)),rep(FALSE,(nrow(fftloess))-as.integer(nrow(fftloess)*degradStartPercent)))
fftloess.df = data.frame(fftloess, rul = rul)
fftloess.hex = as.h2o(fftloess.df)
fftloess.dl = h2o.deeplearning(x = 1:(ncol(fftloess)),y = ncol(fftloess.df),training_frame = fftloess.hex)
if(scale){
center = attr(fftloess,"scaled:center")
scaleBy = attr(fftloess,"scaled:scale")
return(list(model = fftloess.dl,groupSize = groupSize, scale = scale, center = center, scaleBy = scaleBy))
}
return(list(model = fftloess.dl,groupSize = groupSize, scale = scale))
}
#' Create a deep learning model of smoothed stft data with degradation starting at a chosen percentage to predict a chosen number of output states
#' @title Create Multi-State Deep Learning Model (createModel.h2o.logic)
#' @param data The time series data to be transformed and used to train the model
#' @param groupSize The size of the groups that will be fft'd
#' @param levelDegradVec The percentage of the data to pass until the system changes to the lower state
#' @return A deep learning model of the data to predict a state
#' @export
createModel.h2o.levels <- function(data,groupSize,levelDegradVec,scale){
fftloess = get.fftloess(data,groupSize,scale)
if(scale){
center = attr(fftloess,"scaled:center")
scaleBy = attr(fftloess,"scaled:scale")
}
rul = rep(1,nrow(fftloess))
lb = 1
for(i in 1:(length(levelDegradVec))){
ub = as.integer(levelDegradVec[(i)] * (nrow(fftloess)))
rul[lb:ub] = length(levelDegradVec) - (i-2)
lb = ub
}
rul = as.character(rul)
fftloess.df = data.frame(unname(fftloess), rul = rul)
fftloess.hex = as.h2o(fftloess.df)
fftloess.dl = h2o.deeplearning(x = 1:(ncol(fftloess)),y = ncol(fftloess.df),training_frame = fftloess.hex)
if(scale)
return(list(model = fftloess.dl,groupSize = groupSize, scale = scale, center = center, scaleBy = scaleBy))
return(list(model = fftloess.dl,groupSize = groupSize, scale = scale))
}
#' Predict the dependent variable based on the inputted test data and trained model
#' @title Predict STFT (predict.h2o.stft)
#' @param data The test data
#' @param model The trained model either returned from 'createModel.h2o.logic' or 'createModel.h2o.levels' functions
#' @return The predicted output based on the trianed model and test data
#' @export
predict.h2o.stft <- function(data, model){
groupSize = model$groupSize
scale = model$scale
testfft = get.fftloess(data,groupSize,FALSE)
if(scale){
center = model$center
scaleBy = model$scaleBy
registerDoMC()
testfft = foreach(i = 1:groupSize, .combine = 'cbind',.inorder = TRUE) %dopar% {
(testfft[,i]-center[i])/scaleBy[i]
}
testfft = unname(testfft)
# for(i in 1:groupSize)
# testfft[,i] = (testfft[,i]-center[i])/scaleBy[i]
}
testfft.df = data.frame(testfft)
testfft.hex = as.h2o(testfft.df)
predictions = h2o.predict(object = model$model,newdata = testfft.hex)
predictions = as.data.frame(predictions)
return(predictions$predict)
}
#' @export
status.h2o.logic <- function(dataH,dataV,modelH,modelV,groupSize,n){
if(length(predH) > (n*groupSize)){
up = length(predH)
lo = length(predH)-n*groupSize+1
dataH = dataH[lo:up]
dataV = dataV[lo:up]
}
predH = predict.h2o.logic(dataH,groupSize,modelH)
predV = predict.h2o.logic(dataV,groupSize,modelV)
if(all(predH == TRUE)){
if(all(predV == FALSE)){
return(-1)
}
else{
return(1)
}
}
else if(all(predH == FALSE)){
return(-1)
}
else{
if(all(predV == TRUE)){
return(1)
}
else if(all(predV == FALSE)){
return(-1)
}
else{
return(0)
}
}
}
#' @export
predict.status.h2o.logic <- function(predictH,predictV,n){
boundSeq = c(1:(n-1),seq(from=n,to = (length(predictH)), by = n))
status = numeric()
for(i in boundSeq){
lo = numeric()
if(i < n)
lo = 1
else
lo = i - n + 1
predH = predictH[lo:i]
predV = predictV[lo:i]
if(all(predH == TRUE)){
if(all(predV == FALSE)){
status = c(status,-1)
}
else{
status = c(status,1)
}
}
else if(all(predH == FALSE)){
status = c(status,-1)
}
else{
if(all(predV == TRUE)){
status = c(status,1)
}
else if(all(predV == FALSE)){
status = c(status,-1)
}
else{
status = c(status,0)
}
}
}
return(status)
}
#' @export
selfRUL.h2o.fftloess <- function(data,groupSize,degradStartPercent){
fftloess = get.fftloess(data,groupSize)
rul = c(rep(nrow(fftloess),as.integer(nrow(fftloess)*degradStartPercent)),seq(from=(nrow(fftloess)), to = 1, length.out = (nrow(fftloess))-as.integer(nrow(fftloess)*degradStartPercent)))
fftloess.df = data.frame(fftloess,rul = rul)
fftloess.hex = as.h2o(fftloess.df)
fftloess.dl = h2o.deeplearning(x = 1:(ncol(fftloess)),y = ncol(fftloess.df),training_frame = fftloess.hex)
predictions <- h2o.predict(fftloess.dl, fftloess.hex)
predictions = as.data.frame(predictions)
plot.ts(predictions$predict)
}
#' @export
selfRUL.h2o.fftloess.logic <- function(data,groupSize,degradStartPercent){
fftloess = get.fftloess(data,groupSize)
rul = c(rep(TRUE,as.integer(nrow(fftloess)*degradStartPercent)),rep(FALSE,(nrow(fftloess))-as.integer(nrow(fftloess)*degradStartPercent)))
fftloess.df = data.frame(fftloess,rul = rul)
fftloess.hex = as.h2o(fftloess.df)
fftloess.dl = h2o.deeplearning(x = 1:(ncol(fftloess)),y = ncol(fftloess.df),training_frame = fftloess.hex)
predictions <- h2o.predict(fftloess.dl, fftloess.hex)
predictions = as.data.frame(predictions)
plot.ts(predictions$predict)
}
#' @export
selfRUL.h2o.fftloess.logic.skip <- function(data,groupSize,degradStartPercent,smoothSampleNum){
fftloess = get.fftloess.skip(data,groupSize,smoothSampleNum)
#rul = c(rep(nrow(fftloess),as.integer(nrow(fftloess)*degradStartPercent)),seq(from=(nrow(fftloess)), to = 1, length.out = (nrow(fftloess))-as.integer(nrow(fftloess)*degradStartPercent)))
rul = c(rep(TRUE,as.integer(nrow(fftloess)*degradStartPercent)),rep(FALSE,(nrow(fftloess))-as.integer(nrow(fftloess)*degradStartPercent)))
fftloess.df = data.frame(fftloess,rul = rul)
fftloess.hex = as.h2o(fftloess.df)
fftloess.dl = h2o.deeplearning(x = 1:(ncol(fftloess)-1),y = ncol(fftloess),training_frame = fftloess.hex)
predictions <- h2o.predict(fftloess.dl, fftloess.hex)
predictions = as.data.frame(predictions)
plot.ts(predictions$predict)
}
#' @export
plot.fftloess <- function(data, groupSize){
fft1 = abs(fft(data[1:groupSize]))
fft3 = abs(fft(data[(length(data)-groupSize+1):(length(data))]))
if(groupSize%%2 == 0){
fft2 = abs(fft(data[(length(data)/2 - groupSize/2 + 1):(length(data)/2 + groupSize/2)]))
}
else{
fft2 = abs(fft(data[(length(data)/2 - groupSize/2 + 0.5):(length(data)/2) + groupSize/2 - 0.5]))
}
time = 1:groupSize
lo1 = predict(loess(fft1~time))
lo2 = predict(loess(fft2~time))
lo3 = predict(loess(fft3~time))
plot.ts(fft1)
lines(lo1,col= "red")
lines(lo2,col= "blue")
lines(lo3,col= "green")
}
#' @export
status.logic <- function(pred.logic){
greenEnd = (which(pred.logic==FALSE)[1])-1
yellowEnd = greenEnd+1+which(rev(pred.logic)==TRUE)[1]
status = list(greenEnd = greenEnd, yellowEnd = yellowEnd)
return(status)
}
#' @export
plot.data.status <- function(data,status,groupSize,ylab){
plot.ts(data[1:((status$greenEnd)*groupSize)], xlim=c(0,length(data)), ylim=c(min(data),max(data)),col="green",xlab="Time(s)",ylab=ylab)
lines(y=data[(((status$greenEnd)*groupSize)+1):((status$yellowEnd)*groupSize)],x=(((status$greenEnd)*groupSize)+1):((status$yellowEnd)*groupSize), col="yellow")
lines(y=data[(((status$yellowEnd)*groupSize)+1):length(data)],x=(((status$yellowEnd)*groupSize)+1):length(data), col="red")
}
#' #' @export
#' saveModel.fftloess <- function(data.path){
#'
#' }
#'
#' #' @export
#' predict.all.logic <- function(data.path,model.path){
#' load(paste(data.path,"bearingDataH1_1.Rd",sep=""))
#' h1_1=horizontal
#' h1_1=as.numeric(unlist(h1_1))
#' load(paste(data.path,"bearingDataH1_2.Rd",sep=""))
#' h1_2=horizontal
#' h1_2=as.numeric(unlist(h1_2))
#' load(paste(data.path,"bearingDataH1_3.Rd",sep=""))
#' h1_3=horizontal
#' h1_3=as.numeric(unlist(h1_3))
# load(paste(data.path,"bearingDataH1_4.Rd",sep=""))
# h1_4=horizontal
# h1_4=as.numeric(unlist(h1_4))
#' load(paste(data.path,"bearingDataH1_5.Rd",sep=""))
#' h1_5=horizontal
#' h1_5=as.numeric(unlist(h1_5))
#' load(paste(data.path,"bearingDataH1_6.Rd",sep=""))
#' h1_6=horizontal
#' h1_6=as.numeric(unlist(h1_6))
#' load(paste(data.path,"bearingDataH1_7.Rd",sep=""))
#' h1_7=horizontal
#' h1_7=as.numeric(unlist(h1_7))
#' load(paste(data.path,"bearingDataH2_1.Rd",sep=""))
#' h2_1=horizontal
#' h2_1=as.numeric(unlist(h1_1))
#' load(paste(data.path,"bearingDataH2_2.Rd",sep=""))
#' h2_2=horizontal
#' h2_2=as.numeric(unlist(h1_2))
#' load(paste(data.path,"bearingDataH2_3.Rd",sep=""))
#' h2_3=horizontal
#' h2_3=as.numeric(unlist(h2_3))
#' load(paste(data.path,"bearingDataH2_4.Rd",sep=""))
#' h2_4=horizontal
#' h2_4=as.numeric(unlist(h2_4))
#' load(paste(data.path,"bearingDataH2_5.Rd",sep=""))
#' h2_5=horizontal
#' h2_5=as.numeric(unlist(h2_5))
#' load(paste(data.path,"bearingDataH2_6.Rd",sep=""))
#' h2_6=horizontal
#' h2_6=as.numeric(unlist(h2_6))
#' load(paste(data.path,"bearingDataH2_7.Rd",sep=""))
#' h2_7=horizontal
#' h2_7=as.numeric(unlist(h2_7))
#' load(paste(data.path,"bearingDataH3_1.Rd",sep=""))
#' h3_1=horizontal
#' h3_1=as.numeric(unlist(h3_1))
#' load(paste(data.path,"bearingDataH3_2.Rd",sep=""))
#' h3_2=horizontal
#' h3_2=as.numeric(unlist(h3_2))
#' load(paste(data.path,"bearingDataH3_3.Rd",sep=""))
#' h3_3=horizontal
#' h3_3=as.numeric(unlist(h3_3))
#' load(paste(data.path,"bearingDataV1_1.Rd",sep=""))
#' v1_1=vertical
#' v1_1=as.numeric(unlist(v1_1))
#' load(paste(data.path,"bearingDataV1_2.Rd",sep=""))
#' v1_2=horizontal
#' v1_2=as.numeric(unlist(v1_2))
#' load(paste(data.path,"bearingDataV1_3.Rd",sep=""))
#' v1_3=vertical
#' v1_3=as.numeric(unlist(v1_3))
#' load(paste(data.path,"bearingDataV1_4.Rd",sep=""))
#' v1_4=vertical
#' v1_4=as.numeric(unlist(v1_4))
#' load(paste(data.path,"bearingDataV1_5.Rd",sep=""))
#' v1_5=vertical
#' v1_5=as.numeric(unlist(v1_5))
#' load(paste(data.path,"bearingDataV1_6.Rd",sep=""))
#' v1_6=vertical
#' v1_6=as.numeric(unlist(v1_6))
#' load(paste(data.path,"bearingDataV1_7.Rd",sep=""))
#' v1_7=vertical
#' v1_7=as.numeric(unlist(v1_7))
#' load(paste(data.path,"bearingDataV2_1.Rd",sep=""))
#' v2_1=vertical
#' v2_1=as.numeric(unlist(v2_1))
#' load(paste(data.path,"bearingDataV2_2.Rd",sep=""))
#' v2_2=horizontal
#' v2_2=as.numeric(unlist(v2_2))
#' load(paste(data.path,"bearingDataV2_3.Rd",sep=""))
#' v2_3=vertical
#' v2_3=as.numeric(unlist(v2_3))
#' load(paste(data.path,"bearingDataV2_4.Rd",sep=""))
#' v2_4=vertical
#' v2_4=as.numeric(unlist(v2_4))
#' load(paste(data.path,"bearingDataV2_5.Rd",sep=""))
#' v2_5=vertical
#' v2_5=as.numeric(unlist(v2_5))
#' load(paste(data.path,"bearingDataV2_6.Rd",sep=""))
#' v2_6=vertical
#' v2_6=as.numeric(unlist(v2_6))
#' load(paste(data.path,"bearingDataV2_7.Rd",sep=""))
#' v2_7=vertical
#' v2_7=as.numeric(unlist(v2_7))
#' load(paste(data.path,"bearingDataV1_1.Rd",sep=""))
#' v3_1=vertical
#' v3_1=as.numeric(unlist(v1_1))
#' load(paste(data.path,"bearingDataV1_2.Rd",sep=""))
#' v3_2=horizontal
#' v3_2=as.numeric(unlist(v1_2))
#' load(paste(data.path,"bearingDataV3_3.Rd",sep=""))
#' v3_3=vertical
#' v3_3=as.numeric(unlist(v3_3))
#'
#' modelH1_1 = createModel.h2o.logic(h1_1,2048,0.8)
#' save(modelH1_1, file = "modelH1_1.Rd")
#' modelH1_2 = createModel.h2o.logic(h1_2,2048,0.8)
#' save(modelH1_2, file = "modelH1_2.Rd")
#' modelH1_3 = createModel.h2o.logic(h1_3,2048,0.8)
#' save(modelH1_3, file = "modelH1_3.Rd")
#' modelH1_4 = createModel.h2o.logic(h1_4,2048,0.8)
#' save(modelH1_4, file = "modelH1_4.Rd")
#' modelH1_5 = createModel.h2o.logic(h1_5,2048,0.8)
#' save(modelH1_5, file = "modelH1_5.Rd")
#' modelH1_6 = createModel.h2o.logic(h1_6,2048,0.8)
#' save(modelH1_6, file = "modelH1_6.Rd")
#' modelH1_7 = createModel.h2o.logic(h1_7,2048,0.8)
#' save(modelH1_7, file = "modelH1_7.Rd")
#' modelH2_1 = createModel.h2o.logic(h2_1,2048,0.8)
#' save(modelH2_1, file = "modelH2_1.Rd")
#' modelH2_2 = createModel.h2o.logic(h2_2,2048,0.8)
#' save(modelH2_2, file = "modelH2_2.Rd")
#' modelH2_3 = createModel.h2o.logic(h2_3,2048,0.8)
#' save(modelH2_3, file = "modelH2_3.Rd")
#' modelH2_4 = createModel.h2o.logic(h2_4,2048,0.8)
#' save(modelH2_4, file = "modelH2_4.Rd")
#' modelH2_5 = createModel.h2o.logic(h2_5,2048,0.8)
#' save(modelH2_5, file = "modelH2_5.Rd")
#' modelH2_6 = createModel.h2o.logic(h2_6,2048,0.8)
#' save(modelH2_6, file = "modelH2_6.Rd")
#' modelH2_7 = createModel.h2o.logic(h2_7,2048,0.8)
#' save(modelH2_7, file = "modelH2_7.Rd")
#' modelH3_1 = createModel.h2o.logic(h3_1,2048,0.8)
#' save(modelH3_1, file = "modelH3_1.Rd")
#' modelH3_2 = createModel.h2o.logic(h3_2,2048,0.8)
#' save(modelH3_2, file = "modelH3_2.Rd")
#' modelH3_3 = createModel.h2o.logic(h3_3,2048,0.8)
#' save(modelH3_2, file = "modelH3_2.Rd")
#'
#' modelV1_1 = createModel.h2o.logic(v1_1,2048,0.8)
#' save(modelV1_1, file = "modelV1_1.Rd")
#' modelV1_2 = createModel.h2o.logic(v1_2,2048,0.8)
#' save(modelV1_2, file = "modelV1_2.Rd")
#' modelV1_3 = createModel.h2o.logic(v1_3,2048,0.8)
#' save(modelV1_3, file = "modelV1_3.Rd")
#' modelV1_4 = createModel.h2o.logic(v1_4,2048,0.8)
#' save(modelV1_4, file = "modelV1_4.Rd")
#' modelV1_5 = createModel.h2o.logic(v1_5,2048,0.8)
#' save(modelV1_5, file = "modelV1_5.Rd")
#' modelV1_6 = createModel.h2o.logic(v1_6,2048,0.8)
#' save(modelV1_6, file = "modelV1_6.Rd")
#' modelV1_7 = createModel.h2o.logic(v1_7,2048,0.8)
#' save(modelV1_7, file = "modelV1_7.Rd")
#' modelV2_1 = createModel.h2o.logic(v2_1,2048,0.8)
#' save(modelV2_1, file = "modelV2_1.Rd")
#' modelV2_2 = createModel.h2o.logic(v2_2,2048,0.8)
#' save(modelV2_2, file = "modelV2_2.Rd")
#' modelV2_3 = createModel.h2o.logic(v2_3,2048,0.8)
#' save(modelV2_3, file = "modelV2_3.Rd")
#' modelV2_4 = createModel.h2o.logic(v2_4,2048,0.8)
#' save(modelV2_4, file = "modelV2_4.Rd")
#' modelV2_5 = createModel.h2o.logic(v2_5,2048,0.8)
#' save(modelV2_5, file = "modelV2_5.Rd")
#' modelV2_6 = createModel.h2o.logic(v2_6,2048,0.8)
#' save(modelV2_6, file = "modelV2_6.Rd")
#' modelV2_7 = createModel.h2o.logic(v2_7,2048,0.8)
#' save(modelV2_7, file = "modelV2_7.Rd")
#' modelV3_1 = createModel.h2o.logic(v3_1,2048,0.8)
#' save(modelV3_1, file = "modelV3_1.Rd")
#' modelV3_2 = createModel.h2o.logic(v3_2,2048,0.8)
#' save(modelV3_2, file = "modelV3_2.Rd")
#' modelV3_3 = createModel.h2o.logic(v3_3,2048,0.8)
#' save(modelV3_2, file = "modelV3_2.Rd")
#'
#' predictH1_1 = list()
#' predictH1_2 = list()
#' predictH1_3 = list()
#' predictH1_4 = list()
#' predictH1_5 = list()
#' predictH1_6 = list()
#' predictH1_7 = list()
#'
#' predictH1_1[[1]] = predict.h2o.logic(h1_1,2048,modelH1_1)
#' predictH1_1[[2]] = predict.h2o.logic(h1_2,2048,modelH1_1)
#' predictH1_1[[3]] = predict.h2o.logic(h1_3,2048,modelH1_1)
#' predictH1_1[[4]] = predict.h2o.logic(h1_4,2048,modelH1_1)
#' predictH1_1[[5]] = predict.h2o.logic(h1_5,2048,modelH1_1)
#' predictH1_1[[6]] = predict.h2o.logic(h1_6,2048,modelH1_1)
#' predictH1_1[[7]] = predict.h2o.logic(h1_7,2048,modelH1_1)
#'
#' predictH1_2[[1]] = predict.h2o.logic(h1_1,2048,modelH1_2)
#' predictH1_2[[2]] = predict.h2o.logic(h1_2,2048,modelH1_2)
#' predictH1_2[[3]] = predict.h2o.logic(h1_3,2048,modelH1_2)
#' predictH1_2[[4]] = predict.h2o.logic(h1_4,2048,modelH1_2)
#' predictH1_2[[5]] = predict.h2o.logic(h1_5,2048,modelH1_2)
#' predictH1_2[[6]] = predict.h2o.logic(h1_6,2048,modelH1_2)
#' predictH1_2[[7]] = predict.h2o.logic(h1_7,2048,modelH1_2)
#'
#' predictH1_3[[1]] = predict.h2o.logic(h1_1,2048,modelH1_3)
#' predictH1_3[[2]] = predict.h2o.logic(h1_2,2048,modelH1_3)
#' predictH1_3[[3]] = predict.h2o.logic(h1_3,2048,modelH1_3)
#' predictH1_3[[4]] = predict.h2o.logic(h1_4,2048,modelH1_3)
#' predictH1_3[[5]] = predict.h2o.logic(h1_5,2048,modelH1_3)
#' predictH1_3[[6]] = predict.h2o.logic(h1_6,2048,modelH1_3)
#' predictH1_3[[7]] = predict.h2o.logic(h1_7,2048,modelH1_3)
#'
#' predictH1_4[[1]] = predict.h2o.logic(h1_1,2048,modelH1_4)
#' predictH1_4[[2]] = predict.h2o.logic(h1_2,2048,modelH1_4)
#' predictH1_4[[3]] = predict.h2o.logic(h1_3,2048,modelH1_4)
#' predictH1_4[[4]] = predict.h2o.logic(h1_4,2048,modelH1_4)
#' predictH1_4[[5]] = predict.h2o.logic(h1_5,2048,modelH1_4)
#' predictH1_4[[6]] = predict.h2o.logic(h1_6,2048,modelH1_4)
#' predictH1_4[[7]] = predict.h2o.logic(h1_7,2048,modelH1_4)
#'
#' predictH1_5[[1]] = predict.h2o.logic(h1_1,2048,modelH1_5)
#' predictH1_5[[2]] = predict.h2o.logic(h1_2,2048,modelH1_5)
#' predictH1_5[[3]] = predict.h2o.logic(h1_3,2048,modelH1_5)
#' predictH1_5[[4]] = predict.h2o.logic(h1_4,2048,modelH1_5)
#' predictH1_5[[5]] = predict.h2o.logic(h1_5,2048,modelH1_5)
#' predictH1_5[[6]] = predict.h2o.logic(h1_6,2048,modelH1_5)
#' predictH1_5[[7]] = predict.h2o.logic(h1_7,2048,modelH1_5)
#'
#' predictH1_6[[1]] = predict.h2o.logic(h1_1,2048,modelH1_6)
#' predictH1_6[[2]] = predict.h2o.logic(h1_2,2048,modelH1_6)
#' predictH1_6[[3]] = predict.h2o.logic(h1_3,2048,modelH1_6)
#' predictH1_6[[4]] = predict.h2o.logic(h1_4,2048,modelH1_6)
#' predictH1_6[[5]] = predict.h2o.logic(h1_5,2048,modelH1_6)
#' predictH1_6[[6]] = predict.h2o.logic(h1_6,2048,modelH1_6)
#' predictH1_6[[7]] = predict.h2o.logic(h1_7,2048,modelH1_6)
#'
#' predictH1_7[[1]] = predict.h2o.logic(h1_1,2048,modelH1_7)
#' predictH1_7[[2]] = predict.h2o.logic(h1_2,2048,modelH1_7)
#' predictH1_7[[3]] = predict.h2o.logic(h1_3,2048,modelH1_7)
#' predictH1_7[[4]] = predict.h2o.logic(h1_4,2048,modelH1_7)
#' predictH1_7[[5]] = predict.h2o.logic(h1_5,2048,modelH1_7)
#' predictH1_7[[6]] = predict.h2o.logic(h1_6,2048,modelH1_7)
#' predictH1_7[[7]] = predict.h2o.logic(h1_7,2048,modelH1_7)
#'
#' save(predictH1_1,file = "predictH1_1.Rd")
#' save(predictH1_2,file = "predictH1_2.Rd")
#' save(predictH1_3,file = "predictH1_3.Rd")
#' save(predictH1_1,file = "predictH1_4.Rd")
#' save(predictH1_2,file = "predictH1_5.Rd")
#' save(predictH1_3,file = "predictH1_6.Rd")
#' save(predictH1_1,file = "predictH1_7.Rd")
#'
#' predictH2_1 = list()
#' predictH2_2 = list()
#' predictH2_3 = list()
#' predictH2_4 = list()
#' predictH2_5 = list()
#' predictH2_6 = list()
#' predictH2_7 = list()
#'
#' predictH2_1[[1]] = predict.h2o.logic(h2_1,2048,modelH2_1)
#' predictH2_1[[2]] = predict.h2o.logic(h2_2,2048,modelH2_1)
#' predictH2_1[[3]] = predict.h2o.logic(h2_3,2048,modelH2_1)
#' predictH2_1[[4]] = predict.h2o.logic(h2_4,2048,modelH2_1)
#' predictH2_1[[5]] = predict.h2o.logic(h2_5,2048,modelH2_1)
#' predictH2_1[[6]] = predict.h2o.logic(h2_6,2048,modelH2_1)
#' predictH2_1[[7]] = predict.h2o.logic(h2_7,2048,modelH2_1)
#'
#' predictH2_2[[1]] = predict.h2o.logic(h2_1,2048,modelH2_2)
#' predictH2_2[[2]] = predict.h2o.logic(h2_2,2048,modelH2_2)
#' predictH2_2[[3]] = predict.h2o.logic(h2_3,2048,modelH2_2)
#' predictH2_2[[4]] = predict.h2o.logic(h2_4,2048,modelH2_2)
#' predictH2_2[[5]] = predict.h2o.logic(h2_5,2048,modelH2_2)
#' predictH2_2[[6]] = predict.h2o.logic(h2_6,2048,modelH2_2)
#' predictH2_2[[7]] = predict.h2o.logic(h2_7,2048,modelH2_2)
#'
#' predictH2_3[[1]] = predict.h2o.logic(h2_1,2048,modelH2_3)
#' predictH2_3[[2]] = predict.h2o.logic(h2_2,2048,modelH2_3)
#' predictH2_3[[3]] = predict.h2o.logic(h2_3,2048,modelH2_3)
#' predictH2_3[[4]] = predict.h2o.logic(h2_4,2048,modelH2_3)
#' predictH2_3[[5]] = predict.h2o.logic(h2_5,2048,modelH2_3)
#' predictH2_3[[6]] = predict.h2o.logic(h2_6,2048,modelH2_3)
#' predictH2_3[[7]] = predict.h2o.logic(h2_7,2048,modelH2_3)
#'
#' predictH2_4[[1]] = predict.h2o.logic(h2_1,2048,modelH2_4)
#' predictH2_4[[2]] = predict.h2o.logic(h2_2,2048,modelH2_4)
#' predictH2_4[[3]] = predict.h2o.logic(h2_3,2048,modelH2_4)
#' predictH2_4[[4]] = predict.h2o.logic(h2_4,2048,modelH2_4)
#' predictH2_4[[5]] = predict.h2o.logic(h2_5,2048,modelH2_4)
#' predictH2_4[[6]] = predict.h2o.logic(h2_6,2048,modelH2_4)
#' predictH2_4[[7]] = predict.h2o.logic(h2_7,2048,modelH2_4)
#'
#' predictH2_5[[1]] = predict.h2o.logic(h2_1,2048,modelH2_5)
#' predictH2_5[[2]] = predict.h2o.logic(h2_2,2048,modelH2_5)
#' predictH2_5[[3]] = predict.h2o.logic(h2_3,2048,modelH2_5)
#' predictH2_5[[4]] = predict.h2o.logic(h2_4,2048,modelH2_5)
#' predictH2_5[[5]] = predict.h2o.logic(h2_5,2048,modelH2_5)
#' predictH2_5[[6]] = predict.h2o.logic(h2_6,2048,modelH2_5)
#' predictH2_5[[7]] = predict.h2o.logic(h2_7,2048,modelH2_5)
#'
#' predictH2_6[[1]] = predict.h2o.logic(h2_1,2048,modelH2_6)
#' predictH2_6[[2]] = predict.h2o.logic(h2_2,2048,modelH2_6)
#' predictH2_6[[3]] = predict.h2o.logic(h2_3,2048,modelH2_6)
#' predictH2_6[[4]] = predict.h2o.logic(h2_4,2048,modelH2_6)
#' predictH2_6[[5]] = predict.h2o.logic(h2_5,2048,modelH2_6)
#' predictH2_6[[6]] = predict.h2o.logic(h2_6,2048,modelH2_6)
#' predictH2_6[[7]] = predict.h2o.logic(h2_7,2048,modelH2_6)
#'
#' predictH2_7[[1]] = predict.h2o.logic(h2_1,2048,modelH2_7)
#' predictH2_7[[2]] = predict.h2o.logic(h2_2,2048,modelH2_7)
#' predictH2_7[[3]] = predict.h2o.logic(h2_3,2048,modelH2_7)
#' predictH2_7[[4]] = predict.h2o.logic(h2_4,2048,modelH2_7)
#' predictH2_7[[5]] = predict.h2o.logic(h2_5,2048,modelH2_7)
#' predictH2_7[[6]] = predict.h2o.logic(h2_6,2048,modelH2_7)
#' predictH2_7[[7]] = predict.h2o.logic(h2_7,2048,modelH2_7)
#'
#' save(predictH2_1,file = "predictH2_1.Rd")
#' save(predictH2_2,file = "predictH2_2.Rd")
#' save(predictH2_3,file = "predictH2_3.Rd")
#' save(predictH2_1,file = "predictH2_4.Rd")
#' save(predictH2_2,file = "predictH2_5.Rd")
#' save(predictH2_3,file = "predictH2_6.Rd")
#' save(predictH2_1,file = "predictH2_7.Rd")
#'
#' predictH3_1 = list()
#' predictH3_2 = list()
#' predictH3_3 = list()
#'
#' predictH3_1[[1]] = predict.h2o.logic(h3_1,2048,modelH3_1)
#' predictH3_1[[2]] = predict.h2o.logic(h3_2,2048,modelH3_1)
#' predictH3_1[[3]] = predict.h2o.logic(h3_3,2048,modelH3_1)
#'
#' predictH3_2[[1]] = predict.h2o.logic(h3_1,2048,modelH3_2)
#' predictH3_2[[2]] = predict.h2o.logic(h3_2,2048,modelH3_2)
#' predictH3_2[[3]] = predict.h2o.logic(h3_3,2048,modelH3_2)
#'
#' predictH3_3[[1]] = predict.h2o.logic(h3_1,2048,modelH3_3)
#' predictH3_3[[2]] = predict.h2o.logic(h3_2,2048,modelH3_3)
#' predictH3_3[[3]] = predict.h2o.logic(h3_3,2048,modelH3_3)
#'
#' save(predictH3_1,file = "predictH3_1.Rd")
#' save(predictH3_2,file = "predictH3_2.Rd")
#' save(predictH3_3,file = "predictH3_3.Rd")
#'
#' predictV1_1 = list()
#' predictV1_2 = list()
#' predictV1_3 = list()
#' predictV1_4 = list()
#' predictV1_5 = list()
#' predictV1_6 = list()
#' predictV1_7 = list()
#'
#' predictV1_1[[1]] = predict.h2o.logic(v1_1,2048,modelV1_1)
#' predictV1_1[[2]] = predict.h2o.logic(v1_2,2048,modelV1_1)
#' predictV1_1[[3]] = predict.h2o.logic(v1_3,2048,modelV1_1)
#' predictV1_1[[4]] = predict.h2o.logic(v1_4,2048,modelV1_1)
#' predictV1_1[[5]] = predict.h2o.logic(v1_5,2048,modelV1_1)
#' predictV1_1[[6]] = predict.h2o.logic(v1_6,2048,modelV1_1)
#' predictV1_1[[7]] = predict.h2o.logic(v1_7,2048,modelV1_1)
#'
#' predictV1_2[[1]] = predict.h2o.logic(v1_1,2048,modelV1_2)
#' predictV1_2[[2]] = predict.h2o.logic(v1_2,2048,modelV1_2)
#' predictV1_2[[3]] = predict.h2o.logic(v1_3,2048,modelV1_2)
#' predictV1_2[[4]] = predict.h2o.logic(v1_4,2048,modelV1_2)
#' predictV1_2[[5]] = predict.h2o.logic(v1_5,2048,modelV1_2)
#' predictV1_2[[6]] = predict.h2o.logic(v1_6,2048,modelV1_2)
#' predictV1_2[[7]] = predict.h2o.logic(v1_7,2048,modelV1_2)
#'
#' predictV1_3[[1]] = predict.h2o.logic(v1_1,2048,modelV1_3)
#' predictV1_3[[2]] = predict.h2o.logic(v1_2,2048,modelV1_3)
#' predictV1_3[[3]] = predict.h2o.logic(v1_3,2048,modelV1_3)
#' predictV1_3[[4]] = predict.h2o.logic(v1_4,2048,modelV1_3)
#' predictV1_3[[5]] = predict.h2o.logic(v1_5,2048,modelV1_3)
#' predictV1_3[[6]] = predict.h2o.logic(v1_6,2048,modelV1_3)
#' predictV1_3[[7]] = predict.h2o.logic(v1_7,2048,modelV1_3)
#'
#' predictV1_4[[1]] = predict.h2o.logic(v1_1,2048,modelV1_4)
#' predictV1_4[[2]] = predict.h2o.logic(v1_2,2048,modelV1_4)
#' predictV1_4[[3]] = predict.h2o.logic(v1_3,2048,modelV1_4)
#' predictV1_4[[4]] = predict.h2o.logic(v1_4,2048,modelV1_4)
#' predictV1_4[[5]] = predict.h2o.logic(v1_5,2048,modelV1_4)
#' predictV1_4[[6]] = predict.h2o.logic(v1_6,2048,modelV1_4)
#' predictV1_4[[7]] = predict.h2o.logic(v1_7,2048,modelV1_4)
#'
#' predictV1_5[[1]] = predict.h2o.logic(v1_1,2048,modelV1_5)
#' predictV1_5[[2]] = predict.h2o.logic(v1_2,2048,modelV1_5)
#' predictV1_5[[3]] = predict.h2o.logic(v1_3,2048,modelV1_5)
#' predictV1_5[[4]] = predict.h2o.logic(v1_4,2048,modelV1_5)
#' predictV1_5[[5]] = predict.h2o.logic(v1_5,2048,modelV1_5)
#' predictV1_5[[6]] = predict.h2o.logic(v1_6,2048,modelV1_5)
#' predictV1_5[[7]] = predict.h2o.logic(v1_7,2048,modelV1_5)
#'
#' predictV1_6[[1]] = predict.h2o.logic(v1_1,2048,modelV1_6)
#' predictV1_6[[2]] = predict.h2o.logic(v1_2,2048,modelV1_6)
#' predictV1_6[[3]] = predict.h2o.logic(v1_3,2048,modelV1_6)
#' predictV1_6[[4]] = predict.h2o.logic(v1_4,2048,modelV1_6)
#' predictV1_6[[5]] = predict.h2o.logic(v1_5,2048,modelV1_6)
#' predictV1_6[[6]] = predict.h2o.logic(v1_6,2048,modelV1_6)
#' predictV1_6[[7]] = predict.h2o.logic(v1_7,2048,modelV1_6)
#'
#' predictV1_7[[1]] = predict.h2o.logic(v1_1,2048,modelV1_7)
#' predictV1_7[[2]] = predict.h2o.logic(v1_2,2048,modelV1_7)
#' predictV1_7[[3]] = predict.h2o.logic(v1_3,2048,modelV1_7)
#' predictV1_7[[4]] = predict.h2o.logic(v1_4,2048,modelV1_7)
#' predictV1_7[[5]] = predict.h2o.logic(v1_5,2048,modelV1_7)
#' predictV1_7[[6]] = predict.h2o.logic(v1_6,2048,modelV1_7)
#' predictV1_7[[7]] = predict.h2o.logic(v1_7,2048,modelV1_7)
#'
#' save(predictV1_1,file = "predictV1_1.Rd")
#' save(predictV1_2,file = "predictV1_2.Rd")
#' save(predictV1_3,file = "predictV1_3.Rd")
#' save(predictV1_1,file = "predictV1_4.Rd")
#' save(predictV1_2,file = "predictV1_5.Rd")
#' save(predictV1_3,file = "predictV1_6.Rd")
#' save(predictV1_1,file = "predictV1_7.Rd")
#'
#' predictV2_1 = list()
#' predictV2_2 = list()
#' predictV2_3 = list()
#' predictV2_4 = list()
#' predictV2_5 = list()
#' predictV2_6 = list()
#' predictV2_7 = list()
#'
#' predictV2_1[[1]] = predict.h2o.logic(v2_1,2048,modelV2_1)
#' predictV2_1[[2]] = predict.h2o.logic(v2_2,2048,modelV2_1)
#' predictV2_1[[3]] = predict.h2o.logic(v2_3,2048,modelV2_1)
#' predictV2_1[[4]] = predict.h2o.logic(v2_4,2048,modelV2_1)
#' predictV2_1[[5]] = predict.h2o.logic(v2_5,2048,modelV2_1)
#' predictV2_1[[6]] = predict.h2o.logic(v2_6,2048,modelV2_1)
#' predictV2_1[[7]] = predict.h2o.logic(v2_7,2048,modelV2_1)
#'
#' predictV2_2[[1]] = predict.h2o.logic(v2_1,2048,modelV2_2)
#' predictV2_2[[2]] = predict.h2o.logic(v2_2,2048,modelV2_2)
#' predictV2_2[[3]] = predict.h2o.logic(v2_3,2048,modelV2_2)
#' predictV2_2[[4]] = predict.h2o.logic(v2_4,2048,modelV2_2)
#' predictV2_2[[5]] = predict.h2o.logic(v2_5,2048,modelV2_2)
#' predictV2_2[[6]] = predict.h2o.logic(v2_6,2048,modelV2_2)
#' predictV2_2[[7]] = predict.h2o.logic(v2_7,2048,modelV2_2)
#'
#' predictV2_3[[1]] = predict.h2o.logic(v2_1,2048,modelV2_3)
#' predictV2_3[[2]] = predict.h2o.logic(v2_2,2048,modelV2_3)
#' predictV2_3[[3]] = predict.h2o.logic(v2_3,2048,modelV2_3)
#' predictV2_3[[4]] = predict.h2o.logic(v2_4,2048,modelV2_3)
#' predictV2_3[[5]] = predict.h2o.logic(v2_5,2048,modelV2_3)
#' predictV2_3[[6]] = predict.h2o.logic(v2_6,2048,modelV2_3)
#' predictV2_3[[7]] = predict.h2o.logic(v2_7,2048,modelV2_3)
#'
#' predictV2_4[[1]] = predict.h2o.logic(v2_1,2048,modelV2_4)
#' predictV2_4[[2]] = predict.h2o.logic(v2_2,2048,modelV2_4)
#' predictV2_4[[3]] = predict.h2o.logic(v2_3,2048,modelV2_4)
#' predictV2_4[[4]] = predict.h2o.logic(v2_4,2048,modelV2_4)
#' predictV2_4[[5]] = predict.h2o.logic(v2_5,2048,modelV2_4)
#' predictV2_4[[6]] = predict.h2o.logic(v2_6,2048,modelV2_4)
#' predictV2_4[[7]] = predict.h2o.logic(v2_7,2048,modelV2_4)
#'
#' predictV2_5[[1]] = predict.h2o.logic(v2_1,2048,modelV2_5)
#' predictV2_5[[2]] = predict.h2o.logic(v2_2,2048,modelV2_5)
#' predictV2_5[[3]] = predict.h2o.logic(v2_3,2048,modelV2_5)
#' predictV2_5[[4]] = predict.h2o.logic(v2_4,2048,modelV2_5)
#' predictV2_5[[5]] = predict.h2o.logic(v2_5,2048,modelV2_5)
#' predictV2_5[[6]] = predict.h2o.logic(v2_6,2048,modelV2_5)
#' predictV2_5[[7]] = predict.h2o.logic(v2_7,2048,modelV2_5)
#'
#' predictV2_6[[1]] = predict.h2o.logic(v2_1,2048,modelV2_6)
#' predictV2_6[[2]] = predict.h2o.logic(v2_2,2048,modelV2_6)
#' predictV2_6[[3]] = predict.h2o.logic(v2_3,2048,modelV2_6)
#' predictV2_6[[4]] = predict.h2o.logic(v2_4,2048,modelV2_6)
#' predictV2_6[[5]] = predict.h2o.logic(v2_5,2048,modelV2_6)
#' predictV2_6[[6]] = predict.h2o.logic(v2_6,2048,modelV2_6)
#' predictV2_6[[7]] = predict.h2o.logic(v2_7,2048,modelV2_6)
#'
#' predictV2_7[[1]] = predict.h2o.logic(v2_1,2048,modelV2_7)
#' predictV2_7[[2]] = predict.h2o.logic(v2_2,2048,modelV2_7)
#' predictV2_7[[3]] = predict.h2o.logic(v2_3,2048,modelV2_7)
#' predictV2_7[[4]] = predict.h2o.logic(v2_4,2048,modelV2_7)
#' predictV2_7[[5]] = predict.h2o.logic(v2_5,2048,modelV2_7)
#' predictV2_7[[6]] = predict.h2o.logic(v2_6,2048,modelV2_7)
#' predictV2_7[[7]] = predict.h2o.logic(v2_7,2048,modelV2_7)
#'
#' save(predictV2_1,file = "predictV2_1.Rd")
#' save(predictV2_2,file = "predictV2_2.Rd")
#' save(predictV2_3,file = "predictV2_3.Rd")
#' save(predictV2_1,file = "predictV2_4.Rd")
#' save(predictV2_2,file = "predictV2_5.Rd")
#' save(predictV2_3,file = "predictV2_6.Rd")
#' save(predictV2_1,file = "predictV2_7.Rd")
#'
#' predictV3_1 = list()
#' predictV3_2 = list()
#' predictV3_3 = list()
#'
#' predictV3_1[[1]] = predict.h2o.logic(v3_1,2048,modelV3_1)
#' predictV3_1[[2]] = predict.h2o.logic(v3_2,2048,modelV3_1)
#' predictV3_1[[3]] = predict.h2o.logic(v3_3,2048,modelV3_1)
#'
#' predictV3_2[[1]] = predict.h2o.logic(v3_1,2048,modelV3_2)
#' predictV3_2[[2]] = predict.h2o.logic(v3_2,2048,modelV3_2)
#' predictV3_2[[3]] = predict.h2o.logic(v3_3,2048,modelV3_2)
#'
#' predictV3_3[[1]] = predict.h2o.logic(v3_1,2048,modelV3_3)
#' predictV3_3[[2]] = predict.h2o.logic(v3_2,2048,modelV3_3)
#' predictV3_3[[3]] = predict.h2o.logic(v3_3,2048,modelV3_3)
#'
#' save(predictV3_1,file = "predictV3_1.Rd")
#' save(predictV3_2,file = "predictV3_2.Rd")
#' save(predictV3_3,file = "predictV3_3.Rd")
#' }
#'
#' load("modelH1_1.Rd")
#' load("modelH1_2.Rd")
#' load("modelH1_3.Rd")
#' load("modelH1_4.Rd")
#' load("modelH1_5.Rd")
#' load("modelH1_6.Rd")
#' load("modelH1_7.Rd")
#' load("modelH2_1.Rd")
#' load("modelH2_2.Rd")
#' load("modelH2_3.Rd")
#' load("modelH2_4.Rd")
#' load("modelH2_5.Rd")
#' load("modelH2_6.Rd")
#' load("modelH2_7.Rd")
#' load("modelH3_1.Rd")
#' load("modelH3_2.Rd")
#' load("modelH3_3.Rd")
#' load("modelV1_1.Rd")
#' load("modelV1_2.Rd")
#' load("modelV1_3.Rd")
#' load("modelV1_4.Rd")
#' load("modelV1_5.Rd")
#' load("modelV1_6.Rd")
#' load("modelV1_7.Rd")
#' load("modelV2_1.Rd")
#' load("modelV2_2.Rd")
#' load("modelV2_3.Rd")
#' load("modelV2_4.Rd")
#' load("modelV2_5.Rd")
#' load("modelV2_6.Rd")
#' load("modelV2_7.Rd")
#' load("modelV3_1.Rd")
#' load("modelV3_2.Rd")
#' load("modelV3_3.Rd")
# load("~/projects/rspatiotemp/stftData/fftloessH1_1.Rd")
# fftloessH1_1 = fftloess.df
# load("~/projects/rspatiotemp/stftData/fftloessH1_2.Rd")
# fftloessH1_2 = fftloess.df
# load("~/projects/rspatiotemp/stftData/fftloessH1_3.Rd")
# fftloessH1_3 = fftloess.df
# load("~/projects/rspatiotemp/stftData/fftloessH1_4.Rd")
# fftloessH1_4 = fftloess.df
# load("~/projects/rspatiotemp/stftData/fftloessH1_5.Rd")
# fftloessH1_5 = fftloess.df
# load("~/projects/rspatiotemp/stftData/fftloessH1_6.Rd")
# fftloessH1_6 = fftloess.df
# load("~/projects/rspatiotemp/stftData/fftloessH1_7.Rd")
# fftloessH1_7 = fftloess.df
#
# trainH1_1 = rbind(fftloessH1_2,fftloessH1_3)
# trainH1_1 = rbind(trainH1_1,fftloessH1_4)
# trainH1_1 = rbind(trainH1_1,fftloessH1_5)
# trainH1_1 = rbind(trainH1_1,fftloessH1_6)
# trainH1_1 = rbind(trainH1_1,fftloessH1_7)
#
# trainH1_2 = rbind(fftloessH1_1,fftloessH1_3)
# trainH1_2 = rbind(trainH1_2,fftloessH1_4)
# trainH1_2 = rbind(trainH1_2,fftloessH1_5)
# trainH1_2 = rbind(trainH1_2,fftloessH1_6)
# trainH1_2 = rbind(trainH1_2,fftloessH1_7)
#
# trainH1_3 = rbind(fftloessH1_2,fftloessH1_1)
# trainH1_3 = rbind(trainH1_3,fftloessH1_4)
# trainH1_3 = rbind(trainH1_3,fftloessH1_5)
# trainH1_3 = rbind(trainH1_3,fftloessH1_6)
# trainH1_3 = rbind(trainH1_3,fftloessH1_7)
#
# trainH1_4 = rbind(fftloessH1_2,fftloessH1_3)
# trainH1_4 = rbind(trainH1_4,fftloessH1_1)
# trainH1_4 = rbind(trainH1_4,fftloessH1_5)
# trainH1_4 = rbind(trainH1_4,fftloessH1_6)
# trainH1_4 = rbind(trainH1_4,fftloessH1_7)
#
# trainH1_5 = rbind(fftloessH1_2,fftloessH1_3)
# trainH1_5 = rbind(trainH1_5,fftloessH1_4)
# trainH1_5 = rbind(trainH1_5,fftloessH1_1)
# trainH1_5 = rbind(trainH1_5,fftloessH1_6)
# trainH1_5 = rbind(trainH1_5,fftloessH1_7)
#
# trainH1_6 = rbind(fftloessH1_2,fftloessH1_3)
# trainH1_6 = rbind(trainH1_6,fftloessH1_4)
# trainH1_6 = rbind(trainH1_6,fftloessH1_5)
# trainH1_6 = rbind(trainH1_6,fftloessH1_1)
# trainH1_6 = rbind(trainH1_6,fftloessH1_7)
#
# trainH1_7 = rbind(fftloessH1_2,fftloessH1_3)
# trainH1_7 = rbind(trainH1_7,fftloessH1_4)
# trainH1_7 = rbind(trainH1_7,fftloessH1_5)
# trainH1_7 = rbind(trainH1_7,fftloessH1_6)
# trainH1_7 = rbind(trainH1_7,fftloessH1_1)
#
# trainH1_1 = as.h2o(trainH1_1)
# trainH1_2 = as.h2o(trainH1_2)
# trainH1_3 = as.h2o(trainH1_3)
# trainH1_4 = as.h2o(trainH1_4)
# trainH1_5 = as.h2o(trainH1_5)
# trainH1_6 = as.h2o(trainH1_6)
# trainH1_7 = as.h2o(trainH1_7)
#
# trainH1_1.dl = h2o.deeplearning(x = 1:2048, y = 2049, training_frame = trainH1_1)
# trainH1_2.dl = h2o.deeplearning(x = 1:2048, y = 2049, training_frame = trainH1_2)
# trainH1_3.dl = h2o.deeplearning(x = 1:2048, y = 2049, training_frame = trainH1_3)
# trainH1_4.dl = h2o.deeplearning(x = 1:2048, y = 2049, training_frame = trainH1_4)
# trainH1_5.dl = h2o.deeplearning(x = 1:2048, y = 2049, training_frame = trainH1_5)
# trainH1_6.dl = h2o.deeplearning(x = 1:2048, y = 2049, training_frame = trainH1_6)
# trainH1_7.dl = h2o.deeplearning(x = 1:2048, y = 2049, training_frame = trainH1_7)
#
# predH1_1 = h2o.predict(trainH1_1.dl,as.h2o(fftloessH1_1))
# predH1_1 = as.data.frame(predH1_1)$predict
# save(predH1_1, file = "allpredH1_1.Rd")
#
# predH1_2 = h2o.predict(trainH1_2.dl,as.h2o(fftloessH1_2))
# predH1_2 = as.data.frame(predH1_2)$predict
# save(predH1_2, file = "allpredH1_2.Rd")
#
# predH1_3 = h2o.predict(trainH1_3.dl,as.h2o(fftloessH1_3))
# predH1_3 = as.data.frame(predH1_3)$predict
# save(predH1_3, file = "allpredH1_3.Rd")
#
# predH1_4 = h2o.predict(trainH1_4.dl,as.h2o(fftloessH1_4))
# predH1_4 = as.data.frame(predH1_4)$predict
# save(predH1_4, file = "allpredH1_4.Rd")
#
# predH1_5 = h2o.predict(trainH1_5.dl,as.h2o(fftloessH1_5))
# predH1_5 = as.data.frame(predH1_5)$predict
# save(predH1_5, file = "allpredH1_5.Rd")
#
# predH1_6 = h2o.predict(trainH1_6.dl,as.h2o(fftloessH1_6))
# predH1_6 = as.data.frame(predH1_6)$predict
# save(predH1_6, file = "allpredH1_6.Rd")
#
# predH1_7 = h2o.predict(trainH1_7.dl,as.h2o(fftloessH1_7))
# predH1_7 = as.data.frame(predH1_7)$predict
# save(predH1_7, file = "allpredH1_7.Rd")
#
# fftloessH1_1 = fftloess.rul(h1_1,2048,FALSE,0.8)
# fftloessH1_2 = fftloess.rul(h1_2,2048,FALSE,0.8)
# fftloessH1_3 = fftloess.rul(h1_3,2048,FALSE,0.8)
# fftloessH1_4 = fftloess.rul(h1_4,2048,FALSE,0.8)
# fftloessH1_5 = fftloess.rul(h1_5,2048,FALSE,0.8)
# fftloessH1_6 = fftloess.rul(h1_6,2048,FALSE,0.8)
# fftloessH1_7 = fftloess.rul(h1_7,2048,FALSE,0.8)
# library(h2o)
# library(doMC)
# library(foreach)
# h2o.init()
#
#
# #load data
# data.path = "FEMTO Bearing Data/"
# load(paste(data.path,"bearingDataH1_2.Rd",sep=""))
# trainData = as.numeric(unlist(horizontal))
# load(paste(data.path,"bearingDataH1_4.Rd",sep=""))
# test1_4 = as.numeric(unlist(horizontal))
# load(paste(data.path,"bearingDataH1_7.Rd",sep=""))
# test1_7 = as.numeric(unlist(horizontal))
# #create model from bearing1_2 data not scaled
# modelH1_2_stft_F = createModel.h2o.logic(trainData,2048,0.8,FALSE)
# #create model from bearing1_2 data scaled
# modelH1_2_stft_T = createModel.h2o.logic(trainData,2048,0.8,TRUE)
# #These EOLs are predicted using the bearing1_2 trained model
# #EOL1_4 Not Scale.png
# pred1_4_stft_F = predict.h2o.stft(test1_4,modelH1_2_stft_F)
# #EOL1_4 Scale.png
# pred1_4_stft_T = predict.h2o.stft(test1_4,modelH1_2_stft_T)
# #EOL1_7 Not Scale.png
# pred1_7_stft_F = predict.h2o.stft(test1_7,modelH1_2_stft_F)
# #compute status for all EOL not scale
# status1_4_F = status.logic(pred1_4_stft_F)
# status1_4_T = status.logic(pred1_4_stft_T)
# status1_7_F = status.logic(pred1_4_stft_F)
# #plot graphs of EOL behaviour on data
# plot.data.status(test1_4,status1_4_F,2048,"Bearing1_4 Acceleration")
# plot.data.status(test1_4,status1_4_T,2048,"Bearing1_4 Acceleration")
# plot.data.status(test1_7,status1_7_F,2048,"Bearing1_7 Acceleration")
jcpetkovich/rspatiotemp documentation built on May 18, 2019, 10:26 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
library(h2o)
library(doMC)
library(foreach)
h2o.init()
#' Compute the stft of the inputted data and smooth it using loess regression
#' @title STFT loess regression (get.fftloess)
#' @param data The time series data to be transformed
#' @param groupSize The size of the groups that will be fft'd
#' @param scale TRUE or FALSE whether or not to scale the output
#' @return The smoothed stft of the inputted data
#' @export
get.fftloess <- function(data, groupSize,scale){
boundSeq = seq(from=1,to = (length(data)+1), by = groupSize)
len = length(boundSeq) - 1
time = 1:groupSize
# lowerBound = boundSeq[1]
# upperBound = boundSeq[2] - 1
# fft = abs(fft(data[lowerBound:upperBound]))
# lo = predict(loess(fft~time))
registerDoMC()
lo = foreach(i = 1:len,.combine = 'rbind',.inorder = TRUE) %dopar% {
lowerBound = boundSeq[i]
upperBound = boundSeq[i+1] - 1
fft = abs(fft(data[lowerBound:upperBound]))
predict(loess(fft~time))
}
# for(i in 2:len){
# lowerBound = boundSeq[i]
# upperBound = boundSeq[i+1] - 1
# fft = abs(fft(data[lowerBound:upperBound]))
# lo1 = predict(loess(fft~time))
# lo = rbind(lo,lo1)
# }
if(scale)
return(scale(unname(lo)))
return(unname(lo))
}
#' Compute the stft of the inputted data, smooth it using loess regression, but only take a set number of stft pieces
#' @title pieces of STFT loess regression (get.fftloess)
#' @param data The time series data to be transformed
#' @param groupSize The size of the groups that will be fft'd
#' @param smoothSampleNum The number of stft pieces to take
#' @return The smoothed stft of the inputted data
#' @export
get.fftloess.skip <- function(data, groupSize,smoothSampleNum){
boundSeq = seq(from=1,to = (length(data)+1), by = groupSize)
len = length(boundSeq) - 1
boundSeq2 = seq(from = 1, to = length(data),length.out = smoothSampleNum)
time = 1:groupSize
lowerBound = boundSeq[1]
upperBound = boundSeq[2] - 1
fft = abs(fft(data[lowerBound:upperBound]))
lo = predict(loess(fft~time))
for(i in 2:len){
lowerBound = boundSeq[i]
upperBound = boundSeq[i+1] - 1
fft = abs(fft(data[lowerBound:upperBound]))
lo1 = predict(loess(fft~time))[boundSeq2]
lo = rbind(lo,lo1)
}
return(lo)
}
#' Create a deep learning model of smoothed stft data with degradation starting at a chosen percentage to predict a logical variable
#' @title Create Logical Deep Learning Model (createModel.h2o.logic)
#' @param data The time series data to be transformed and used to train the model
#' @param groupSize The size of the groups that will be fft'd
#' @param degradStartPercent The percentage of the data to pass until the system begins to degrade
#' @return A deep learning model of the data to predict a logical output
#' @export
createModel.h2o.logic <- function(data,groupSize,degradStartPercent,scale){
fftloess = get.fftloess(data,groupSize,scale)
rul = c(rep(TRUE,as.integer(nrow(fftloess)*degradStartPercent)),rep(FALSE,(nrow(fftloess))-as.integer(nrow(fftloess)*degradStartPercent)))
fftloess.df = data.frame(fftloess, rul = rul)
fftloess.hex = as.h2o(fftloess.df)
fftloess.dl = h2o.deeplearning(x = 1:(ncol(fftloess)),y = ncol(fftloess.df),training_frame = fftloess.hex)
if(scale){
center = attr(fftloess,"scaled:center")
scaleBy = attr(fftloess,"scaled:scale")
return(list(model = fftloess.dl,groupSize = groupSize, scale = scale, center = center, scaleBy = scaleBy))
}
return(list(model = fftloess.dl,groupSize = groupSize, scale = scale))
}
#' Create a deep learning model of smoothed stft data with degradation starting at a chosen percentage to predict a chosen number of output states
#' @title Create Multi-State Deep Learning Model (createModel.h2o.logic)
#' @param data The time series data to be transformed and used to train the model
#' @param groupSize The size of the groups that will be fft'd
#' @param levelDegradVec The percentage of the data to pass until the system changes to the lower state
#' @return A deep learning model of the data to predict a state
#' @export
createModel.h2o.levels <- function(data,groupSize,levelDegradVec,scale){
fftloess = get.fftloess(data,groupSize,scale)
if(scale){
center = attr(fftloess,"scaled:center")
scaleBy = attr(fftloess,"scaled:scale")
}
rul = rep(1,nrow(fftloess))
lb = 1
for(i in 1:(length(levelDegradVec))){
ub = as.integer(levelDegradVec[(i)] * (nrow(fftloess)))
rul[lb:ub] = length(levelDegradVec) - (i-2)
lb = ub
}
rul = as.character(rul)
fftloess.df = data.frame(unname(fftloess), rul = rul)
fftloess.hex = as.h2o(fftloess.df)
fftloess.dl = h2o.deeplearning(x = 1:(ncol(fftloess)),y = ncol(fftloess.df),training_frame = fftloess.hex)
if(scale)
return(list(model = fftloess.dl,groupSize = groupSize, scale = scale, center = center, scaleBy = scaleBy))
return(list(model = fftloess.dl,groupSize = groupSize, scale = scale))
}
#' Predict the dependent variable based on the inputted test data and trained model
#' @title Predict STFT (predict.h2o.stft)
#' @param data The test data
#' @param model The trained model either returned from 'createModel.h2o.logic' or 'createModel.h2o.levels' functions
#' @return The predicted output based on the trianed model and test data
#' @export
predict.h2o.stft <- function(data, model){
groupSize = model$groupSize
scale = model$scale
testfft = get.fftloess(data,groupSize,FALSE)
if(scale){
center = model$center
scaleBy = model$scaleBy
registerDoMC()
testfft = foreach(i = 1:groupSize, .combine = 'cbind',.inorder = TRUE) %dopar% {
(testfft[,i]-center[i])/scaleBy[i]
}
testfft = unname(testfft)
# for(i in 1:groupSize)
# testfft[,i] = (testfft[,i]-center[i])/scaleBy[i]
}
testfft.df = data.frame(testfft)
testfft.hex = as.h2o(testfft.df)
predictions = h2o.predict(object = model$model,newdata = testfft.hex)
predictions = as.data.frame(predictions)
return(predictions$predict)
}
#' @export
status.h2o.logic <- function(dataH,dataV,modelH,modelV,groupSize,n){
if(length(predH) > (n*groupSize)){
up = length(predH)
lo = length(predH)-n*groupSize+1
dataH = dataH[lo:up]
dataV = dataV[lo:up]
}
predH = predict.h2o.logic(dataH,groupSize,modelH)
predV = predict.h2o.logic(dataV,groupSize,modelV)
if(all(predH == TRUE)){
if(all(predV == FALSE)){
return(-1)
}
else{
return(1)
}
}
else if(all(predH == FALSE)){
return(-1)
}
else{
if(all(predV == TRUE)){
return(1)
}
else if(all(predV == FALSE)){
return(-1)
}
else{
return(0)
}
}
}
#' @export
predict.status.h2o.logic <- function(predictH,predictV,n){
boundSeq = c(1:(n-1),seq(from=n,to = (length(predictH)), by = n))
status = numeric()
for(i in boundSeq){
lo = numeric()
if(i < n)
lo = 1
else
lo = i - n + 1
predH = predictH[lo:i]
predV = predictV[lo:i]
if(all(predH == TRUE)){
if(all(predV == FALSE)){
status = c(status,-1)
}
else{
status = c(status,1)
}
}
else if(all(predH == FALSE)){
status = c(status,-1)
}
else{
if(all(predV == TRUE)){
status = c(status,1)
}
else if(all(predV == FALSE)){
status = c(status,-1)
}
else{
status = c(status,0)
}
}
}
return(status)
}
#' @export
selfRUL.h2o.fftloess <- function(data,groupSize,degradStartPercent){
fftloess = get.fftloess(data,groupSize)
rul = c(rep(nrow(fftloess),as.integer(nrow(fftloess)*degradStartPercent)),seq(from=(nrow(fftloess)), to = 1, length.out = (nrow(fftloess))-as.integer(nrow(fftloess)*degradStartPercent)))
fftloess.df = data.frame(fftloess,rul = rul)
fftloess.hex = as.h2o(fftloess.df)
fftloess.dl = h2o.deeplearning(x = 1:(ncol(fftloess)),y = ncol(fftloess.df),training_frame = fftloess.hex)
predictions <- h2o.predict(fftloess.dl, fftloess.hex)
predictions = as.data.frame(predictions)
plot.ts(predictions$predict)
}
#' @export
selfRUL.h2o.fftloess.logic <- function(data,groupSize,degradStartPercent){
fftloess = get.fftloess(data,groupSize)
rul = c(rep(TRUE,as.integer(nrow(fftloess)*degradStartPercent)),rep(FALSE,(nrow(fftloess))-as.integer(nrow(fftloess)*degradStartPercent)))
fftloess.df = data.frame(fftloess,rul = rul)
fftloess.hex = as.h2o(fftloess.df)
fftloess.dl = h2o.deeplearning(x = 1:(ncol(fftloess)),y = ncol(fftloess.df),training_frame = fftloess.hex)
predictions <- h2o.predict(fftloess.dl, fftloess.hex)
predictions = as.data.frame(predictions)
plot.ts(predictions$predict)
}
#' @export
selfRUL.h2o.fftloess.logic.skip <- function(data,groupSize,degradStartPercent,smoothSampleNum){
fftloess = get.fftloess.skip(data,groupSize,smoothSampleNum)
#rul = c(rep(nrow(fftloess),as.integer(nrow(fftloess)*degradStartPercent)),seq(from=(nrow(fftloess)), to = 1, length.out = (nrow(fftloess))-as.integer(nrow(fftloess)*degradStartPercent)))
rul = c(rep(TRUE,as.integer(nrow(fftloess)*degradStartPercent)),rep(FALSE,(nrow(fftloess))-as.integer(nrow(fftloess)*degradStartPercent)))
fftloess.df = data.frame(fftloess,rul = rul)
fftloess.hex = as.h2o(fftloess.df)
fftloess.dl = h2o.deeplearning(x = 1:(ncol(fftloess)-1),y = ncol(fftloess),training_frame = fftloess.hex)
predictions <- h2o.predict(fftloess.dl, fftloess.hex)
predictions = as.data.frame(predictions)
plot.ts(predictions$predict)
}
#' @export
plot.fftloess <- function(data, groupSize){
fft1 = abs(fft(data[1:groupSize]))
fft3 = abs(fft(data[(length(data)-groupSize+1):(length(data))]))
if(groupSize%%2 == 0){
fft2 = abs(fft(data[(length(data)/2 - groupSize/2 + 1):(length(data)/2 + groupSize/2)]))
}
else{
fft2 = abs(fft(data[(length(data)/2 - groupSize/2 + 0.5):(length(data)/2) + groupSize/2 - 0.5]))
}
time = 1:groupSize
lo1 = predict(loess(fft1~time))
lo2 = predict(loess(fft2~time))
lo3 = predict(loess(fft3~time))
plot.ts(fft1)
lines(lo1,col= "red")
lines(lo2,col= "blue")
lines(lo3,col= "green")
}
#' @export
status.logic <- function(pred.logic){
greenEnd = (which(pred.logic==FALSE)[1])-1
yellowEnd = greenEnd+1+which(rev(pred.logic)==TRUE)[1]
status = list(greenEnd = greenEnd, yellowEnd = yellowEnd)
return(status)
}
#' @export
plot.data.status <- function(data,status,groupSize,ylab){
plot.ts(data[1:((status$greenEnd)*groupSize)], xlim=c(0,length(data)), ylim=c(min(data),max(data)),col="green",xlab="Time(s)",ylab=ylab)
lines(y=data[(((status$greenEnd)*groupSize)+1):((status$yellowEnd)*groupSize)],x=(((status$greenEnd)*groupSize)+1):((status$yellowEnd)*groupSize), col="yellow")
lines(y=data[(((status$yellowEnd)*groupSize)+1):length(data)],x=(((status$yellowEnd)*groupSize)+1):length(data), col="red")
}
#' #' @export
#' saveModel.fftloess <- function(data.path){
#'
#' }
#'
#' #' @export
#' predict.all.logic <- function(data.path,model.path){
#' load(paste(data.path,"bearingDataH1_1.Rd",sep=""))
#' h1_1=horizontal
#' h1_1=as.numeric(unlist(h1_1))
#' load(paste(data.path,"bearingDataH1_2.Rd",sep=""))
#' h1_2=horizontal
#' h1_2=as.numeric(unlist(h1_2))
#' load(paste(data.path,"bearingDataH1_3.Rd",sep=""))
#' h1_3=horizontal
#' h1_3=as.numeric(unlist(h1_3))
# load(paste(data.path,"bearingDataH1_4.Rd",sep=""))
# h1_4=horizontal
# h1_4=as.numeric(unlist(h1_4))
#' load(paste(data.path,"bearingDataH1_5.Rd",sep=""))
#' h1_5=horizontal
#' h1_5=as.numeric(unlist(h1_5))
#' load(paste(data.path,"bearingDataH1_6.Rd",sep=""))
#' h1_6=horizontal
#' h1_6=as.numeric(unlist(h1_6))
#' load(paste(data.path,"bearingDataH1_7.Rd",sep=""))
#' h1_7=horizontal
#' h1_7=as.numeric(unlist(h1_7))
#' load(paste(data.path,"bearingDataH2_1.Rd",sep=""))
#' h2_1=horizontal
#' h2_1=as.numeric(unlist(h1_1))
#' load(paste(data.path,"bearingDataH2_2.Rd",sep=""))
#' h2_2=horizontal
#' h2_2=as.numeric(unlist(h1_2))
#' load(paste(data.path,"bearingDataH2_3.Rd",sep=""))
#' h2_3=horizontal
#' h2_3=as.numeric(unlist(h2_3))
#' load(paste(data.path,"bearingDataH2_4.Rd",sep=""))
#' h2_4=horizontal
#' h2_4=as.numeric(unlist(h2_4))
#' load(paste(data.path,"bearingDataH2_5.Rd",sep=""))
#' h2_5=horizontal
#' h2_5=as.numeric(unlist(h2_5))
#' load(paste(data.path,"bearingDataH2_6.Rd",sep=""))
#' h2_6=horizontal
#' h2_6=as.numeric(unlist(h2_6))
#' load(paste(data.path,"bearingDataH2_7.Rd",sep=""))
#' h2_7=horizontal
#' h2_7=as.numeric(unlist(h2_7))
#' load(paste(data.path,"bearingDataH3_1.Rd",sep=""))
#' h3_1=horizontal
#' h3_1=as.numeric(unlist(h3_1))
#' load(paste(data.path,"bearingDataH3_2.Rd",sep=""))
#' h3_2=horizontal
#' h3_2=as.numeric(unlist(h3_2))
#' load(paste(data.path,"bearingDataH3_3.Rd",sep=""))
#' h3_3=horizontal
#' h3_3=as.numeric(unlist(h3_3))
#' load(paste(data.path,"bearingDataV1_1.Rd",sep=""))
#' v1_1=vertical
#' v1_1=as.numeric(unlist(v1_1))
#' load(paste(data.path,"bearingDataV1_2.Rd",sep=""))
#' v1_2=horizontal
#' v1_2=as.numeric(unlist(v1_2))
#' load(paste(data.path,"bearingDataV1_3.Rd",sep=""))
#' v1_3=vertical
#' v1_3=as.numeric(unlist(v1_3))
#' load(paste(data.path,"bearingDataV1_4.Rd",sep=""))
#' v1_4=vertical
#' v1_4=as.numeric(unlist(v1_4))
#' load(paste(data.path,"bearingDataV1_5.Rd",sep=""))
#' v1_5=vertical
#' v1_5=as.numeric(unlist(v1_5))
#' load(paste(data.path,"bearingDataV1_6.Rd",sep=""))
#' v1_6=vertical
#' v1_6=as.numeric(unlist(v1_6))
#' load(paste(data.path,"bearingDataV1_7.Rd",sep=""))
#' v1_7=vertical
#' v1_7=as.numeric(unlist(v1_7))
#' load(paste(data.path,"bearingDataV2_1.Rd",sep=""))
#' v2_1=vertical
#' v2_1=as.numeric(unlist(v2_1))
#' load(paste(data.path,"bearingDataV2_2.Rd",sep=""))
#' v2_2=horizontal
#' v2_2=as.numeric(unlist(v2_2))
#' load(paste(data.path,"bearingDataV2_3.Rd",sep=""))
#' v2_3=vertical
#' v2_3=as.numeric(unlist(v2_3))
#' load(paste(data.path,"bearingDataV2_4.Rd",sep=""))
#' v2_4=vertical
#' v2_4=as.numeric(unlist(v2_4))
#' load(paste(data.path,"bearingDataV2_5.Rd",sep=""))
#' v2_5=vertical
#' v2_5=as.numeric(unlist(v2_5))
#' load(paste(data.path,"bearingDataV2_6.Rd",sep=""))
#' v2_6=vertical
#' v2_6=as.numeric(unlist(v2_6))
#' load(paste(data.path,"bearingDataV2_7.Rd",sep=""))
#' v2_7=vertical
#' v2_7=as.numeric(unlist(v2_7))
#' load(paste(data.path,"bearingDataV1_1.Rd",sep=""))
#' v3_1=vertical
#' v3_1=as.numeric(unlist(v1_1))
#' load(paste(data.path,"bearingDataV1_2.Rd",sep=""))
#' v3_2=horizontal
#' v3_2=as.numeric(unlist(v1_2))
#' load(paste(data.path,"bearingDataV3_3.Rd",sep=""))
#' v3_3=vertical
#' v3_3=as.numeric(unlist(v3_3))
#'
#' modelH1_1 = createModel.h2o.logic(h1_1,2048,0.8)
#' save(modelH1_1, file = "modelH1_1.Rd")
#' modelH1_2 = createModel.h2o.logic(h1_2,2048,0.8)
#' save(modelH1_2, file = "modelH1_2.Rd")
#' modelH1_3 = createModel.h2o.logic(h1_3,2048,0.8)
#' save(modelH1_3, file = "modelH1_3.Rd")
#' modelH1_4 = createModel.h2o.logic(h1_4,2048,0.8)
#' save(modelH1_4, file = "modelH1_4.Rd")
#' modelH1_5 = createModel.h2o.logic(h1_5,2048,0.8)
#' save(modelH1_5, file = "modelH1_5.Rd")
#' modelH1_6 = createModel.h2o.logic(h1_6,2048,0.8)
#' save(modelH1_6, file = "modelH1_6.Rd")
#' modelH1_7 = createModel.h2o.logic(h1_7,2048,0.8)
#' save(modelH1_7, file = "modelH1_7.Rd")
#' modelH2_1 = createModel.h2o.logic(h2_1,2048,0.8)
#' save(modelH2_1, file = "modelH2_1.Rd")
#' modelH2_2 = createModel.h2o.logic(h2_2,2048,0.8)
#' save(modelH2_2, file = "modelH2_2.Rd")
#' modelH2_3 = createModel.h2o.logic(h2_3,2048,0.8)
#' save(modelH2_3, file = "modelH2_3.Rd")
#' modelH2_4 = createModel.h2o.logic(h2_4,2048,0.8)
#' save(modelH2_4, file = "modelH2_4.Rd")
#' modelH2_5 = createModel.h2o.logic(h2_5,2048,0.8)
#' save(modelH2_5, file = "modelH2_5.Rd")
#' modelH2_6 = createModel.h2o.logic(h2_6,2048,0.8)
#' save(modelH2_6, file = "modelH2_6.Rd")
#' modelH2_7 = createModel.h2o.logic(h2_7,2048,0.8)
#' save(modelH2_7, file = "modelH2_7.Rd")
#' modelH3_1 = createModel.h2o.logic(h3_1,2048,0.8)
#' save(modelH3_1, file = "modelH3_1.Rd")
#' modelH3_2 = createModel.h2o.logic(h3_2,2048,0.8)
#' save(modelH3_2, file = "modelH3_2.Rd")
#' modelH3_3 = createModel.h2o.logic(h3_3,2048,0.8)
#' save(modelH3_2, file = "modelH3_2.Rd")
#'
#' modelV1_1 = createModel.h2o.logic(v1_1,2048,0.8)
#' save(modelV1_1, file = "modelV1_1.Rd")
#' modelV1_2 = createModel.h2o.logic(v1_2,2048,0.8)
#' save(modelV1_2, file = "modelV1_2.Rd")
#' modelV1_3 = createModel.h2o.logic(v1_3,2048,0.8)
#' save(modelV1_3, file = "modelV1_3.Rd")
#' modelV1_4 = createModel.h2o.logic(v1_4,2048,0.8)
#' save(modelV1_4, file = "modelV1_4.Rd")
#' modelV1_5 = createModel.h2o.logic(v1_5,2048,0.8)
#' save(modelV1_5, file = "modelV1_5.Rd")
#' modelV1_6 = createModel.h2o.logic(v1_6,2048,0.8)
#' save(modelV1_6, file = "modelV1_6.Rd")
#' modelV1_7 = createModel.h2o.logic(v1_7,2048,0.8)
#' save(modelV1_7, file = "modelV1_7.Rd")
#' modelV2_1 = createModel.h2o.logic(v2_1,2048,0.8)
#' save(modelV2_1, file = "modelV2_1.Rd")
#' modelV2_2 = createModel.h2o.logic(v2_2,2048,0.8)
#' save(modelV2_2, file = "modelV2_2.Rd")
#' modelV2_3 = createModel.h2o.logic(v2_3,2048,0.8)
#' save(modelV2_3, file = "modelV2_3.Rd")
#' modelV2_4 = createModel.h2o.logic(v2_4,2048,0.8)
#' save(modelV2_4, file = "modelV2_4.Rd")
#' modelV2_5 = createModel.h2o.logic(v2_5,2048,0.8)
#' save(modelV2_5, file = "modelV2_5.Rd")
#' modelV2_6 = createModel.h2o.logic(v2_6,2048,0.8)
#' save(modelV2_6, file = "modelV2_6.Rd")
#' modelV2_7 = createModel.h2o.logic(v2_7,2048,0.8)
#' save(modelV2_7, file = "modelV2_7.Rd")
#' modelV3_1 = createModel.h2o.logic(v3_1,2048,0.8)
#' save(modelV3_1, file = "modelV3_1.Rd")
#' modelV3_2 = createModel.h2o.logic(v3_2,2048,0.8)
#' save(modelV3_2, file = "modelV3_2.Rd")
#' modelV3_3 = createModel.h2o.logic(v3_3,2048,0.8)
#' save(modelV3_2, file = "modelV3_2.Rd")
#'
#' predictH1_1 = list()
#' predictH1_2 = list()
#' predictH1_3 = list()
#' predictH1_4 = list()
#' predictH1_5 = list()
#' predictH1_6 = list()
#' predictH1_7 = list()
#'
#' predictH1_1[[1]] = predict.h2o.logic(h1_1,2048,modelH1_1)
#' predictH1_1[[2]] = predict.h2o.logic(h1_2,2048,modelH1_1)
#' predictH1_1[[3]] = predict.h2o.logic(h1_3,2048,modelH1_1)
#' predictH1_1[[4]] = predict.h2o.logic(h1_4,2048,modelH1_1)
#' predictH1_1[[5]] = predict.h2o.logic(h1_5,2048,modelH1_1)
#' predictH1_1[[6]] = predict.h2o.logic(h1_6,2048,modelH1_1)
#' predictH1_1[[7]] = predict.h2o.logic(h1_7,2048,modelH1_1)
#'
#' predictH1_2[[1]] = predict.h2o.logic(h1_1,2048,modelH1_2)
#' predictH1_2[[2]] = predict.h2o.logic(h1_2,2048,modelH1_2)
#' predictH1_2[[3]] = predict.h2o.logic(h1_3,2048,modelH1_2)
#' predictH1_2[[4]] = predict.h2o.logic(h1_4,2048,modelH1_2)
#' predictH1_2[[5]] = predict.h2o.logic(h1_5,2048,modelH1_2)
#' predictH1_2[[6]] = predict.h2o.logic(h1_6,2048,modelH1_2)
#' predictH1_2[[7]] = predict.h2o.logic(h1_7,2048,modelH1_2)
#'
#' predictH1_3[[1]] = predict.h2o.logic(h1_1,2048,modelH1_3)
#' predictH1_3[[2]] = predict.h2o.logic(h1_2,2048,modelH1_3)
#' predictH1_3[[3]] = predict.h2o.logic(h1_3,2048,modelH1_3)
#' predictH1_3[[4]] = predict.h2o.logic(h1_4,2048,modelH1_3)
#' predictH1_3[[5]] = predict.h2o.logic(h1_5,2048,modelH1_3)
#' predictH1_3[[6]] = predict.h2o.logic(h1_6,2048,modelH1_3)
#' predictH1_3[[7]] = predict.h2o.logic(h1_7,2048,modelH1_3)
#'
#' predictH1_4[[1]] = predict.h2o.logic(h1_1,2048,modelH1_4)
#' predictH1_4[[2]] = predict.h2o.logic(h1_2,2048,modelH1_4)
#' predictH1_4[[3]] = predict.h2o.logic(h1_3,2048,modelH1_4)
#' predictH1_4[[4]] = predict.h2o.logic(h1_4,2048,modelH1_4)
#' predictH1_4[[5]] = predict.h2o.logic(h1_5,2048,modelH1_4)
#' predictH1_4[[6]] = predict.h2o.logic(h1_6,2048,modelH1_4)
#' predictH1_4[[7]] = predict.h2o.logic(h1_7,2048,modelH1_4)
#'
#' predictH1_5[[1]] = predict.h2o.logic(h1_1,2048,modelH1_5)
#' predictH1_5[[2]] = predict.h2o.logic(h1_2,2048,modelH1_5)
#' predictH1_5[[3]] = predict.h2o.logic(h1_3,2048,modelH1_5)
#' predictH1_5[[4]] = predict.h2o.logic(h1_4,2048,modelH1_5)
#' predictH1_5[[5]] = predict.h2o.logic(h1_5,2048,modelH1_5)
#' predictH1_5[[6]] = predict.h2o.logic(h1_6,2048,modelH1_5)
#' predictH1_5[[7]] = predict.h2o.logic(h1_7,2048,modelH1_5)
#'
#' predictH1_6[[1]] = predict.h2o.logic(h1_1,2048,modelH1_6)
#' predictH1_6[[2]] = predict.h2o.logic(h1_2,2048,modelH1_6)
#' predictH1_6[[3]] = predict.h2o.logic(h1_3,2048,modelH1_6)
#' predictH1_6[[4]] = predict.h2o.logic(h1_4,2048,modelH1_6)
#' predictH1_6[[5]] = predict.h2o.logic(h1_5,2048,modelH1_6)
#' predictH1_6[[6]] = predict.h2o.logic(h1_6,2048,modelH1_6)
#' predictH1_6[[7]] = predict.h2o.logic(h1_7,2048,modelH1_6)
#'
#' predictH1_7[[1]] = predict.h2o.logic(h1_1,2048,modelH1_7)
#' predictH1_7[[2]] = predict.h2o.logic(h1_2,2048,modelH1_7)
#' predictH1_7[[3]] = predict.h2o.logic(h1_3,2048,modelH1_7)
#' predictH1_7[[4]] = predict.h2o.logic(h1_4,2048,modelH1_7)
#' predictH1_7[[5]] = predict.h2o.logic(h1_5,2048,modelH1_7)
#' predictH1_7[[6]] = predict.h2o.logic(h1_6,2048,modelH1_7)
#' predictH1_7[[7]] = predict.h2o.logic(h1_7,2048,modelH1_7)
#'
#' save(predictH1_1,file = "predictH1_1.Rd")
#' save(predictH1_2,file = "predictH1_2.Rd")
#' save(predictH1_3,file = "predictH1_3.Rd")
#' save(predictH1_1,file = "predictH1_4.Rd")
#' save(predictH1_2,file = "predictH1_5.Rd")
#' save(predictH1_3,file = "predictH1_6.Rd")
#' save(predictH1_1,file = "predictH1_7.Rd")
#'
#' predictH2_1 = list()
#' predictH2_2 = list()
#' predictH2_3 = list()
#' predictH2_4 = list()
#' predictH2_5 = list()
#' predictH2_6 = list()
#' predictH2_7 = list()
#'
#' predictH2_1[[1]] = predict.h2o.logic(h2_1,2048,modelH2_1)
#' predictH2_1[[2]] = predict.h2o.logic(h2_2,2048,modelH2_1)
#' predictH2_1[[3]] = predict.h2o.logic(h2_3,2048,modelH2_1)
#' predictH2_1[[4]] = predict.h2o.logic(h2_4,2048,modelH2_1)
#' predictH2_1[[5]] = predict.h2o.logic(h2_5,2048,modelH2_1)
#' predictH2_1[[6]] = predict.h2o.logic(h2_6,2048,modelH2_1)
#' predictH2_1[[7]] = predict.h2o.logic(h2_7,2048,modelH2_1)
#'
#' predictH2_2[[1]] = predict.h2o.logic(h2_1,2048,modelH2_2)
#' predictH2_2[[2]] = predict.h2o.logic(h2_2,2048,modelH2_2)
#' predictH2_2[[3]] = predict.h2o.logic(h2_3,2048,modelH2_2)
#' predictH2_2[[4]] = predict.h2o.logic(h2_4,2048,modelH2_2)
#' predictH2_2[[5]] = predict.h2o.logic(h2_5,2048,modelH2_2)
#' predictH2_2[[6]] = predict.h2o.logic(h2_6,2048,modelH2_2)
#' predictH2_2[[7]] = predict.h2o.logic(h2_7,2048,modelH2_2)
#'
#' predictH2_3[[1]] = predict.h2o.logic(h2_1,2048,modelH2_3)
#' predictH2_3[[2]] = predict.h2o.logic(h2_2,2048,modelH2_3)
#' predictH2_3[[3]] = predict.h2o.logic(h2_3,2048,modelH2_3)
#' predictH2_3[[4]] = predict.h2o.logic(h2_4,2048,modelH2_3)
#' predictH2_3[[5]] = predict.h2o.logic(h2_5,2048,modelH2_3)
#' predictH2_3[[6]] = predict.h2o.logic(h2_6,2048,modelH2_3)
#' predictH2_3[[7]] = predict.h2o.logic(h2_7,2048,modelH2_3)
#'
#' predictH2_4[[1]] = predict.h2o.logic(h2_1,2048,modelH2_4)
#' predictH2_4[[2]] = predict.h2o.logic(h2_2,2048,modelH2_4)
#' predictH2_4[[3]] = predict.h2o.logic(h2_3,2048,modelH2_4)
#' predictH2_4[[4]] = predict.h2o.logic(h2_4,2048,modelH2_4)
#' predictH2_4[[5]] = predict.h2o.logic(h2_5,2048,modelH2_4)
#' predictH2_4[[6]] = predict.h2o.logic(h2_6,2048,modelH2_4)
#' predictH2_4[[7]] = predict.h2o.logic(h2_7,2048,modelH2_4)
#'
#' predictH2_5[[1]] = predict.h2o.logic(h2_1,2048,modelH2_5)
#' predictH2_5[[2]] = predict.h2o.logic(h2_2,2048,modelH2_5)
#' predictH2_5[[3]] = predict.h2o.logic(h2_3,2048,modelH2_5)
#' predictH2_5[[4]] = predict.h2o.logic(h2_4,2048,modelH2_5)
#' predictH2_5[[5]] = predict.h2o.logic(h2_5,2048,modelH2_5)
#' predictH2_5[[6]] = predict.h2o.logic(h2_6,2048,modelH2_5)
#' predictH2_5[[7]] = predict.h2o.logic(h2_7,2048,modelH2_5)
#'
#' predictH2_6[[1]] = predict.h2o.logic(h2_1,2048,modelH2_6)
#' predictH2_6[[2]] = predict.h2o.logic(h2_2,2048,modelH2_6)
#' predictH2_6[[3]] = predict.h2o.logic(h2_3,2048,modelH2_6)
#' predictH2_6[[4]] = predict.h2o.logic(h2_4,2048,modelH2_6)
#' predictH2_6[[5]] = predict.h2o.logic(h2_5,2048,modelH2_6)
#' predictH2_6[[6]] = predict.h2o.logic(h2_6,2048,modelH2_6)
#' predictH2_6[[7]] = predict.h2o.logic(h2_7,2048,modelH2_6)
#'
#' predictH2_7[[1]] = predict.h2o.logic(h2_1,2048,modelH2_7)
#' predictH2_7[[2]] = predict.h2o.logic(h2_2,2048,modelH2_7)
#' predictH2_7[[3]] = predict.h2o.logic(h2_3,2048,modelH2_7)
#' predictH2_7[[4]] = predict.h2o.logic(h2_4,2048,modelH2_7)
#' predictH2_7[[5]] = predict.h2o.logic(h2_5,2048,modelH2_7)
#' predictH2_7[[6]] = predict.h2o.logic(h2_6,2048,modelH2_7)
#' predictH2_7[[7]] = predict.h2o.logic(h2_7,2048,modelH2_7)
#'
#' save(predictH2_1,file = "predictH2_1.Rd")
#' save(predictH2_2,file = "predictH2_2.Rd")
#' save(predictH2_3,file = "predictH2_3.Rd")
#' save(predictH2_1,file = "predictH2_4.Rd")
#' save(predictH2_2,file = "predictH2_5.Rd")
#' save(predictH2_3,file = "predictH2_6.Rd")
#' save(predictH2_1,file = "predictH2_7.Rd")
#'
#' predictH3_1 = list()
#' predictH3_2 = list()
#' predictH3_3 = list()
#'
#' predictH3_1[[1]] = predict.h2o.logic(h3_1,2048,modelH3_1)
#' predictH3_1[[2]] = predict.h2o.logic(h3_2,2048,modelH3_1)
#' predictH3_1[[3]] = predict.h2o.logic(h3_3,2048,modelH3_1)
#'
#' predictH3_2[[1]] = predict.h2o.logic(h3_1,2048,modelH3_2)
#' predictH3_2[[2]] = predict.h2o.logic(h3_2,2048,modelH3_2)
#' predictH3_2[[3]] = predict.h2o.logic(h3_3,2048,modelH3_2)
#'
#' predictH3_3[[1]] = predict.h2o.logic(h3_1,2048,modelH3_3)
#' predictH3_3[[2]] = predict.h2o.logic(h3_2,2048,modelH3_3)
#' predictH3_3[[3]] = predict.h2o.logic(h3_3,2048,modelH3_3)
#'
#' save(predictH3_1,file = "predictH3_1.Rd")
#' save(predictH3_2,file = "predictH3_2.Rd")
#' save(predictH3_3,file = "predictH3_3.Rd")
#'
#' predictV1_1 = list()
#' predictV1_2 = list()
#' predictV1_3 = list()
#' predictV1_4 = list()
#' predictV1_5 = list()
#' predictV1_6 = list()
#' predictV1_7 = list()
#'
#' predictV1_1[[1]] = predict.h2o.logic(v1_1,2048,modelV1_1)
#' predictV1_1[[2]] = predict.h2o.logic(v1_2,2048,modelV1_1)
#' predictV1_1[[3]] = predict.h2o.logic(v1_3,2048,modelV1_1)
#' predictV1_1[[4]] = predict.h2o.logic(v1_4,2048,modelV1_1)
#' predictV1_1[[5]] = predict.h2o.logic(v1_5,2048,modelV1_1)
#' predictV1_1[[6]] = predict.h2o.logic(v1_6,2048,modelV1_1)
#' predictV1_1[[7]] = predict.h2o.logic(v1_7,2048,modelV1_1)
#'
#' predictV1_2[[1]] = predict.h2o.logic(v1_1,2048,modelV1_2)
#' predictV1_2[[2]] = predict.h2o.logic(v1_2,2048,modelV1_2)
#' predictV1_2[[3]] = predict.h2o.logic(v1_3,2048,modelV1_2)
#' predictV1_2[[4]] = predict.h2o.logic(v1_4,2048,modelV1_2)
#' predictV1_2[[5]] = predict.h2o.logic(v1_5,2048,modelV1_2)
#' predictV1_2[[6]] = predict.h2o.logic(v1_6,2048,modelV1_2)
#' predictV1_2[[7]] = predict.h2o.logic(v1_7,2048,modelV1_2)
#'
#' predictV1_3[[1]] = predict.h2o.logic(v1_1,2048,modelV1_3)
#' predictV1_3[[2]] = predict.h2o.logic(v1_2,2048,modelV1_3)
#' predictV1_3[[3]] = predict.h2o.logic(v1_3,2048,modelV1_3)
#' predictV1_3[[4]] = predict.h2o.logic(v1_4,2048,modelV1_3)
#' predictV1_3[[5]] = predict.h2o.logic(v1_5,2048,modelV1_3)
#' predictV1_3[[6]] = predict.h2o.logic(v1_6,2048,modelV1_3)
#' predictV1_3[[7]] = predict.h2o.logic(v1_7,2048,modelV1_3)
#'
#' predictV1_4[[1]] = predict.h2o.logic(v1_1,2048,modelV1_4)
#' predictV1_4[[2]] = predict.h2o.logic(v1_2,2048,modelV1_4)
#' predictV1_4[[3]] = predict.h2o.logic(v1_3,2048,modelV1_4)
#' predictV1_4[[4]] = predict.h2o.logic(v1_4,2048,modelV1_4)
#' predictV1_4[[5]] = predict.h2o.logic(v1_5,2048,modelV1_4)
#' predictV1_4[[6]] = predict.h2o.logic(v1_6,2048,modelV1_4)
#' predictV1_4[[7]] = predict.h2o.logic(v1_7,2048,modelV1_4)
#'
#' predictV1_5[[1]] = predict.h2o.logic(v1_1,2048,modelV1_5)
#' predictV1_5[[2]] = predict.h2o.logic(v1_2,2048,modelV1_5)
#' predictV1_5[[3]] = predict.h2o.logic(v1_3,2048,modelV1_5)
#' predictV1_5[[4]] = predict.h2o.logic(v1_4,2048,modelV1_5)
#' predictV1_5[[5]] = predict.h2o.logic(v1_5,2048,modelV1_5)
#' predictV1_5[[6]] = predict.h2o.logic(v1_6,2048,modelV1_5)
#' predictV1_5[[7]] = predict.h2o.logic(v1_7,2048,modelV1_5)
#'
#' predictV1_6[[1]] = predict.h2o.logic(v1_1,2048,modelV1_6)
#' predictV1_6[[2]] = predict.h2o.logic(v1_2,2048,modelV1_6)
#' predictV1_6[[3]] = predict.h2o.logic(v1_3,2048,modelV1_6)
#' predictV1_6[[4]] = predict.h2o.logic(v1_4,2048,modelV1_6)
#' predictV1_6[[5]] = predict.h2o.logic(v1_5,2048,modelV1_6)
#' predictV1_6[[6]] = predict.h2o.logic(v1_6,2048,modelV1_6)
#' predictV1_6[[7]] = predict.h2o.logic(v1_7,2048,modelV1_6)
#'
#' predictV1_7[[1]] = predict.h2o.logic(v1_1,2048,modelV1_7)
#' predictV1_7[[2]] = predict.h2o.logic(v1_2,2048,modelV1_7)
#' predictV1_7[[3]] = predict.h2o.logic(v1_3,2048,modelV1_7)
#' predictV1_7[[4]] = predict.h2o.logic(v1_4,2048,modelV1_7)
#' predictV1_7[[5]] = predict.h2o.logic(v1_5,2048,modelV1_7)
#' predictV1_7[[6]] = predict.h2o.logic(v1_6,2048,modelV1_7)
#' predictV1_7[[7]] = predict.h2o.logic(v1_7,2048,modelV1_7)
#'
#' save(predictV1_1,file = "predictV1_1.Rd")
#' save(predictV1_2,file = "predictV1_2.Rd")
#' save(predictV1_3,file = "predictV1_3.Rd")
#' save(predictV1_1,file = "predictV1_4.Rd")
#' save(predictV1_2,file = "predictV1_5.Rd")
#' save(predictV1_3,file = "predictV1_6.Rd")
#' save(predictV1_1,file = "predictV1_7.Rd")
#'
#' predictV2_1 = list()
#' predictV2_2 = list()
#' predictV2_3 = list()
#' predictV2_4 = list()
#' predictV2_5 = list()
#' predictV2_6 = list()
#' predictV2_7 = list()
#'
#' predictV2_1[[1]] = predict.h2o.logic(v2_1,2048,modelV2_1)
#' predictV2_1[[2]] = predict.h2o.logic(v2_2,2048,modelV2_1)
#' predictV2_1[[3]] = predict.h2o.logic(v2_3,2048,modelV2_1)
#' predictV2_1[[4]] = predict.h2o.logic(v2_4,2048,modelV2_1)
#' predictV2_1[[5]] = predict.h2o.logic(v2_5,2048,modelV2_1)
#' predictV2_1[[6]] = predict.h2o.logic(v2_6,2048,modelV2_1)
#' predictV2_1[[7]] = predict.h2o.logic(v2_7,2048,modelV2_1)
#'
#' predictV2_2[[1]] = predict.h2o.logic(v2_1,2048,modelV2_2)
#' predictV2_2[[2]] = predict.h2o.logic(v2_2,2048,modelV2_2)
#' predictV2_2[[3]] = predict.h2o.logic(v2_3,2048,modelV2_2)
#' predictV2_2[[4]] = predict.h2o.logic(v2_4,2048,modelV2_2)
#' predictV2_2[[5]] = predict.h2o.logic(v2_5,2048,modelV2_2)
#' predictV2_2[[6]] = predict.h2o.logic(v2_6,2048,modelV2_2)
#' predictV2_2[[7]] = predict.h2o.logic(v2_7,2048,modelV2_2)
#'
#' predictV2_3[[1]] = predict.h2o.logic(v2_1,2048,modelV2_3)
#' predictV2_3[[2]] = predict.h2o.logic(v2_2,2048,modelV2_3)
#' predictV2_3[[3]] = predict.h2o.logic(v2_3,2048,modelV2_3)
#' predictV2_3[[4]] = predict.h2o.logic(v2_4,2048,modelV2_3)
#' predictV2_3[[5]] = predict.h2o.logic(v2_5,2048,modelV2_3)
#' predictV2_3[[6]] = predict.h2o.logic(v2_6,2048,modelV2_3)
#' predictV2_3[[7]] = predict.h2o.logic(v2_7,2048,modelV2_3)
#'
#' predictV2_4[[1]] = predict.h2o.logic(v2_1,2048,modelV2_4)
#' predictV2_4[[2]] = predict.h2o.logic(v2_2,2048,modelV2_4)
#' predictV2_4[[3]] = predict.h2o.logic(v2_3,2048,modelV2_4)
#' predictV2_4[[4]] = predict.h2o.logic(v2_4,2048,modelV2_4)
#' predictV2_4[[5]] = predict.h2o.logic(v2_5,2048,modelV2_4)
#' predictV2_4[[6]] = predict.h2o.logic(v2_6,2048,modelV2_4)
#' predictV2_4[[7]] = predict.h2o.logic(v2_7,2048,modelV2_4)
#'
#' predictV2_5[[1]] = predict.h2o.logic(v2_1,2048,modelV2_5)
#' predictV2_5[[2]] = predict.h2o.logic(v2_2,2048,modelV2_5)
#' predictV2_5[[3]] = predict.h2o.logic(v2_3,2048,modelV2_5)
#' predictV2_5[[4]] = predict.h2o.logic(v2_4,2048,modelV2_5)
#' predictV2_5[[5]] = predict.h2o.logic(v2_5,2048,modelV2_5)
#' predictV2_5[[6]] = predict.h2o.logic(v2_6,2048,modelV2_5)
#' predictV2_5[[7]] = predict.h2o.logic(v2_7,2048,modelV2_5)
#'
#' predictV2_6[[1]] = predict.h2o.logic(v2_1,2048,modelV2_6)
#' predictV2_6[[2]] = predict.h2o.logic(v2_2,2048,modelV2_6)
#' predictV2_6[[3]] = predict.h2o.logic(v2_3,2048,modelV2_6)
#' predictV2_6[[4]] = predict.h2o.logic(v2_4,2048,modelV2_6)
#' predictV2_6[[5]] = predict.h2o.logic(v2_5,2048,modelV2_6)
#' predictV2_6[[6]] = predict.h2o.logic(v2_6,2048,modelV2_6)
#' predictV2_6[[7]] = predict.h2o.logic(v2_7,2048,modelV2_6)
#'
#' predictV2_7[[1]] = predict.h2o.logic(v2_1,2048,modelV2_7)
#' predictV2_7[[2]] = predict.h2o.logic(v2_2,2048,modelV2_7)
#' predictV2_7[[3]] = predict.h2o.logic(v2_3,2048,modelV2_7)
#' predictV2_7[[4]] = predict.h2o.logic(v2_4,2048,modelV2_7)
#' predictV2_7[[5]] = predict.h2o.logic(v2_5,2048,modelV2_7)
#' predictV2_7[[6]] = predict.h2o.logic(v2_6,2048,modelV2_7)
#' predictV2_7[[7]] = predict.h2o.logic(v2_7,2048,modelV2_7)
#'
#' save(predictV2_1,file = "predictV2_1.Rd")
#' save(predictV2_2,file = "predictV2_2.Rd")
#' save(predictV2_3,file = "predictV2_3.Rd")
#' save(predictV2_1,file = "predictV2_4.Rd")
#' save(predictV2_2,file = "predictV2_5.Rd")
#' save(predictV2_3,file = "predictV2_6.Rd")
#' save(predictV2_1,file = "predictV2_7.Rd")
#'
#' predictV3_1 = list()
#' predictV3_2 = list()
#' predictV3_3 = list()
#'
#' predictV3_1[[1]] = predict.h2o.logic(v3_1,2048,modelV3_1)
#' predictV3_1[[2]] = predict.h2o.logic(v3_2,2048,modelV3_1)
#' predictV3_1[[3]] = predict.h2o.logic(v3_3,2048,modelV3_1)
#'
#' predictV3_2[[1]] = predict.h2o.logic(v3_1,2048,modelV3_2)
#' predictV3_2[[2]] = predict.h2o.logic(v3_2,2048,modelV3_2)
#' predictV3_2[[3]] = predict.h2o.logic(v3_3,2048,modelV3_2)
#'
#' predictV3_3[[1]] = predict.h2o.logic(v3_1,2048,modelV3_3)
#' predictV3_3[[2]] = predict.h2o.logic(v3_2,2048,modelV3_3)
#' predictV3_3[[3]] = predict.h2o.logic(v3_3,2048,modelV3_3)
#'
#' save(predictV3_1,file = "predictV3_1.Rd")
#' save(predictV3_2,file = "predictV3_2.Rd")
#' save(predictV3_3,file = "predictV3_3.Rd")
#' }
#'
#' load("modelH1_1.Rd")
#' load("modelH1_2.Rd")
#' load("modelH1_3.Rd")
#' load("modelH1_4.Rd")
#' load("modelH1_5.Rd")
#' load("modelH1_6.Rd")
#' load("modelH1_7.Rd")
#' load("modelH2_1.Rd")
#' load("modelH2_2.Rd")
#' load("modelH2_3.Rd")
#' load("modelH2_4.Rd")
#' load("modelH2_5.Rd")
#' load("modelH2_6.Rd")
#' load("modelH2_7.Rd")
#' load("modelH3_1.Rd")
#' load("modelH3_2.Rd")
#' load("modelH3_3.Rd")
#' load("modelV1_1.Rd")
#' load("modelV1_2.Rd")
#' load("modelV1_3.Rd")
#' load("modelV1_4.Rd")
#' load("modelV1_5.Rd")
#' load("modelV1_6.Rd")
#' load("modelV1_7.Rd")
#' load("modelV2_1.Rd")
#' load("modelV2_2.Rd")
#' load("modelV2_3.Rd")
#' load("modelV2_4.Rd")
#' load("modelV2_5.Rd")
#' load("modelV2_6.Rd")
#' load("modelV2_7.Rd")
#' load("modelV3_1.Rd")
#' load("modelV3_2.Rd")
#' load("modelV3_3.Rd")
# load("~/projects/rspatiotemp/stftData/fftloessH1_1.Rd")
# fftloessH1_1 = fftloess.df
# load("~/projects/rspatiotemp/stftData/fftloessH1_2.Rd")
# fftloessH1_2 = fftloess.df
# load("~/projects/rspatiotemp/stftData/fftloessH1_3.Rd")
# fftloessH1_3 = fftloess.df
# load("~/projects/rspatiotemp/stftData/fftloessH1_4.Rd")
# fftloessH1_4 = fftloess.df
# load("~/projects/rspatiotemp/stftData/fftloessH1_5.Rd")
# fftloessH1_5 = fftloess.df
# load("~/projects/rspatiotemp/stftData/fftloessH1_6.Rd")
# fftloessH1_6 = fftloess.df
# load("~/projects/rspatiotemp/stftData/fftloessH1_7.Rd")
# fftloessH1_7 = fftloess.df
#
# trainH1_1 = rbind(fftloessH1_2,fftloessH1_3)
# trainH1_1 = rbind(trainH1_1,fftloessH1_4)
# trainH1_1 = rbind(trainH1_1,fftloessH1_5)
# trainH1_1 = rbind(trainH1_1,fftloessH1_6)
# trainH1_1 = rbind(trainH1_1,fftloessH1_7)
#
# trainH1_2 = rbind(fftloessH1_1,fftloessH1_3)
# trainH1_2 = rbind(trainH1_2,fftloessH1_4)
# trainH1_2 = rbind(trainH1_2,fftloessH1_5)
# trainH1_2 = rbind(trainH1_2,fftloessH1_6)
# trainH1_2 = rbind(trainH1_2,fftloessH1_7)
#
# trainH1_3 = rbind(fftloessH1_2,fftloessH1_1)
# trainH1_3 = rbind(trainH1_3,fftloessH1_4)
# trainH1_3 = rbind(trainH1_3,fftloessH1_5)
# trainH1_3 = rbind(trainH1_3,fftloessH1_6)
# trainH1_3 = rbind(trainH1_3,fftloessH1_7)
#
# trainH1_4 = rbind(fftloessH1_2,fftloessH1_3)
# trainH1_4 = rbind(trainH1_4,fftloessH1_1)
# trainH1_4 = rbind(trainH1_4,fftloessH1_5)
# trainH1_4 = rbind(trainH1_4,fftloessH1_6)
# trainH1_4 = rbind(trainH1_4,fftloessH1_7)
#
# trainH1_5 = rbind(fftloessH1_2,fftloessH1_3)
# trainH1_5 = rbind(trainH1_5,fftloessH1_4)
# trainH1_5 = rbind(trainH1_5,fftloessH1_1)
# trainH1_5 = rbind(trainH1_5,fftloessH1_6)
# trainH1_5 = rbind(trainH1_5,fftloessH1_7)
#
# trainH1_6 = rbind(fftloessH1_2,fftloessH1_3)
# trainH1_6 = rbind(trainH1_6,fftloessH1_4)
# trainH1_6 = rbind(trainH1_6,fftloessH1_5)
# trainH1_6 = rbind(trainH1_6,fftloessH1_1)
# trainH1_6 = rbind(trainH1_6,fftloessH1_7)
#
# trainH1_7 = rbind(fftloessH1_2,fftloessH1_3)
# trainH1_7 = rbind(trainH1_7,fftloessH1_4)
# trainH1_7 = rbind(trainH1_7,fftloessH1_5)
# trainH1_7 = rbind(trainH1_7,fftloessH1_6)
# trainH1_7 = rbind(trainH1_7,fftloessH1_1)
#
# trainH1_1 = as.h2o(trainH1_1)
# trainH1_2 = as.h2o(trainH1_2)
# trainH1_3 = as.h2o(trainH1_3)
# trainH1_4 = as.h2o(trainH1_4)
# trainH1_5 = as.h2o(trainH1_5)
# trainH1_6 = as.h2o(trainH1_6)
# trainH1_7 = as.h2o(trainH1_7)
#
# trainH1_1.dl = h2o.deeplearning(x = 1:2048, y = 2049, training_frame = trainH1_1)
# trainH1_2.dl = h2o.deeplearning(x = 1:2048, y = 2049, training_frame = trainH1_2)
# trainH1_3.dl = h2o.deeplearning(x = 1:2048, y = 2049, training_frame = trainH1_3)
# trainH1_4.dl = h2o.deeplearning(x = 1:2048, y = 2049, training_frame = trainH1_4)
# trainH1_5.dl = h2o.deeplearning(x = 1:2048, y = 2049, training_frame = trainH1_5)
# trainH1_6.dl = h2o.deeplearning(x = 1:2048, y = 2049, training_frame = trainH1_6)
# trainH1_7.dl = h2o.deeplearning(x = 1:2048, y = 2049, training_frame = trainH1_7)
#
# predH1_1 = h2o.predict(trainH1_1.dl,as.h2o(fftloessH1_1))
# predH1_1 = as.data.frame(predH1_1)$predict
# save(predH1_1, file = "allpredH1_1.Rd")
#
# predH1_2 = h2o.predict(trainH1_2.dl,as.h2o(fftloessH1_2))
# predH1_2 = as.data.frame(predH1_2)$predict
# save(predH1_2, file = "allpredH1_2.Rd")
#
# predH1_3 = h2o.predict(trainH1_3.dl,as.h2o(fftloessH1_3))
# predH1_3 = as.data.frame(predH1_3)$predict
# save(predH1_3, file = "allpredH1_3.Rd")
#
# predH1_4 = h2o.predict(trainH1_4.dl,as.h2o(fftloessH1_4))
# predH1_4 = as.data.frame(predH1_4)$predict
# save(predH1_4, file = "allpredH1_4.Rd")
#
# predH1_5 = h2o.predict(trainH1_5.dl,as.h2o(fftloessH1_5))
# predH1_5 = as.data.frame(predH1_5)$predict
# save(predH1_5, file = "allpredH1_5.Rd")
#
# predH1_6 = h2o.predict(trainH1_6.dl,as.h2o(fftloessH1_6))
# predH1_6 = as.data.frame(predH1_6)$predict
# save(predH1_6, file = "allpredH1_6.Rd")
#
# predH1_7 = h2o.predict(trainH1_7.dl,as.h2o(fftloessH1_7))
# predH1_7 = as.data.frame(predH1_7)$predict
# save(predH1_7, file = "allpredH1_7.Rd")
#
# fftloessH1_1 = fftloess.rul(h1_1,2048,FALSE,0.8)
# fftloessH1_2 = fftloess.rul(h1_2,2048,FALSE,0.8)
# fftloessH1_3 = fftloess.rul(h1_3,2048,FALSE,0.8)
# fftloessH1_4 = fftloess.rul(h1_4,2048,FALSE,0.8)
# fftloessH1_5 = fftloess.rul(h1_5,2048,FALSE,0.8)
# fftloessH1_6 = fftloess.rul(h1_6,2048,FALSE,0.8)
# fftloessH1_7 = fftloess.rul(h1_7,2048,FALSE,0.8)
# library(h2o)
# library(doMC)
# library(foreach)
# h2o.init()
#
#
# #load data
# data.path = "FEMTO Bearing Data/"
# load(paste(data.path,"bearingDataH1_2.Rd",sep=""))
# trainData = as.numeric(unlist(horizontal))
# load(paste(data.path,"bearingDataH1_4.Rd",sep=""))
# test1_4 = as.numeric(unlist(horizontal))
# load(paste(data.path,"bearingDataH1_7.Rd",sep=""))
# test1_7 = as.numeric(unlist(horizontal))
# #create model from bearing1_2 data not scaled
# modelH1_2_stft_F = createModel.h2o.logic(trainData,2048,0.8,FALSE)
# #create model from bearing1_2 data scaled
# modelH1_2_stft_T = createModel.h2o.logic(trainData,2048,0.8,TRUE)
# #These EOLs are predicted using the bearing1_2 trained model
# #EOL1_4 Not Scale.png
# pred1_4_stft_F = predict.h2o.stft(test1_4,modelH1_2_stft_F)
# #EOL1_4 Scale.png
# pred1_4_stft_T = predict.h2o.stft(test1_4,modelH1_2_stft_T)
# #EOL1_7 Not Scale.png
# pred1_7_stft_F = predict.h2o.stft(test1_7,modelH1_2_stft_F)
# #compute status for all EOL not scale
# status1_4_F = status.logic(pred1_4_stft_F)
# status1_4_T = status.logic(pred1_4_stft_T)
# status1_7_F = status.logic(pred1_4_stft_F)
# #plot graphs of EOL behaviour on data
# plot.data.status(test1_4,status1_4_F,2048,"Bearing1_4 Acceleration")
# plot.data.status(test1_4,status1_4_T,2048,"Bearing1_4 Acceleration")
# plot.data.status(test1_7,status1_7_F,2048,"Bearing1_7 Acceleration")
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.