Nothing
R/deepForest.R
In deepdive: Deep Learning for General Purpose
Defines functions
deepforest
Documented in
deepforest
#' @title Build or train bagged deeptree or deepnet of multiple architecture
#' @description Build or train bagged deeptree or deepnet of multiple architecture.Based on error choice either select best model or average multiple model with random variable cut,data cut and architechture
#' @param x a data frame with input variables
#' @param y a data frame with ouptut variable
#' @param networkCount Integer, Number of deepnet or deeptree to build
#' @param layerChoice vector, different layer choices
#' @param unitsChoice vector , number of units choice
#' @param cutVarSizePercent ratio, percentage of variable to for each network
#' @param cutDataSizePercent ratio, percentage of data to for each network
#' @param activation choose from "sigmoid","relu","sin","cos","none".Activations will be randomly chosen from chosen. Default is relu and sin
#' @param reluLeak numeric. Applicable when activation is "relu". Specify value between 0 any number close to zero below 1. Eg: 0.01,0.001 etc
#' @param modelType one of "regress","binary","multiClass". "regress" for regression will create a linear single unit output layer. "binary" will create a single unit sigmoid activated layer. "multiClass" will create layer with units corresponding to number of output classes with softmax activation.
#' @param iterations integer. This indicates number of iteratios or epochs in backpropagtion .The default value is 500.
#' @param eta numeric.Hyperparameter,sets the Learning rate for backpropagation. Eta determines the convergence ability and speed of convergence.
#' @param seed numeric. Set seed with this parameter. Incase of sin activation sometimes changing seed can yeild better results. Default is 2
#' @param gradientClip numeric. Hyperparameter numeric value which limits gradient size for weight update operation in backpropagation. Default is 0.8 . It can take any postive value.
#' @param regularisePar numeric. L2 Regularisation Parameter .
#' @param optimiser one of "gradientDescent","momentum","rmsProp","adam". Default value "adam"
#' @param parMomentum numeric. Applicable for optimiser "mometum" and "adam"
#' @param inputSizeImpact numeric. Adjusts the gradient size by factor of percentage of rows in input. For very small data set setting this to 0 could yeild faster result. Default is 1.
#' @param parRmsPropZeroAdjust numeric. Applicable for optimiser "rmsProp" and "adam"
#' @param parRmsProp numeric.Applicable for optimiser "rmsProp" and "adam"
#' @param treeLeaves vector.Optional , leaves numbers from externally trained tree model can be supplied here. If supplied then model will not build a explicit tree and just fit a neural network to mentioned leaves.
#' @param treeMinSplitPercent numeric. This parameter controls depth of tree setting min split count for leaf subdivision as percentage of observations. Final minimum split will be chosen as max of count calculted with treeMinSplitPercent and treeMinSplitCount. Default 0.3. Range 0 to 1.
#' @param treeMinSplitCount numeric. This parameter controls depth of tree setting min split count.Final minimum split will be chosen as max of count calculted with treeMinSplitPercent and treeMinSplitCount. Default 30
#' @param treeCp complexity parameter. \code{\link{rpart.control}}
#' @param errorCover Ratio. Deault is 0.2 i.e all models within 20 percent error of best model will be selected.
#' @param treeAugment logical. If True fits deeptree and if False fits deepnet. Default is T
#' @param printItrSize numeric. Number of iterations after which progress message should be shown. Default value 100 and for iterations below 100 atleast 5 messages will be seen
#' @param showProgress logical. True will show progress and F will not show progress
#' @param stopError Numeric. Rmse at which iterations can be stopped. Default is 0.01, can be set as NA in case all iterations needs to run.
#' @param miniBatchSize integer. Set the mini batch size for mini batch gradient
#' @param useBatchProgress logical. Applicable for miniBatch , setting T will use show rmse in Batch and F will show error on full dataset. For large dataset set T
#' @return returns model object which can be passed into \code{\link{predict.deepforest}}
#' @export
#' @import rpart
#' @import data.table
#' @importFrom stringr str_remove_all
#' @importFrom treeClust rpart.predict.leaves
#' @importFrom graphics barplot
#' @importFrom stats formula predict runif
#' @examples
#'
#' require(deepdive)
#'
#'x<-data.frame(x1=runif(10),x2=runif(10))
#'y<-data.frame(y=10*x$x1+20*x$x2+20)
#'
#'mdeepf<-deepforest(x,y,
#' networkCount=2,
#' layerChoice=c(2:3),
#' unitsChoice=c(4:10),
#' cutVarSizePercent=0.6,
#' cutDataSizePercent=0.6,
#' activation = c('relu',"sin"),
#' reluLeak=0.01,
#' modelType ='regress',
#' iterations = 10,
#' eta = 10 ^-2,
#' seed=2,
#' gradientClip=0.8,
#' regularisePar=0,
#' optimiser="adam",
#' parMomentum=0.9,
#' inputSizeImpact=1,
#' parRmsPropZeroAdjust=10^-8,
#' parRmsProp=0.9999,
#' treeLeaves=NA,
#' treeMinSplitPercent=0.3,
#' treeMinSplitCount=100,
#' treeCp=0.01 ,
#' errorCover=0.2,
#' treeAugment=TRUE,
#' printItrSize=100,
#' showProgress=TRUE,
#' stopError=0.01,
#' miniBatchSize=64,
#' useBatchProgress=TRUE)
deepforest<-function(x,y,
networkCount=3,
layerChoice=c(2:3),
unitsChoice=c(4:10),
cutVarSizePercent=0.6,
cutDataSizePercent=0.6,
activation = c('sigmoid',"sigmoid"),
reluLeak=0,
modelType ='regress',
iterations = 500,
eta = 10 ^-2,
seed=2,
gradientClip=0.8,
regularisePar=0,
optimiser="adam",
parMomentum=0.9,
inputSizeImpact=1,
parRmsPropZeroAdjust=10^-8,
parRmsProp=0.9999,
treeLeaves=NA,
treeMinSplitPercent=0.3,
treeMinSplitCount=100,
treeCp=0.01 ,
errorCover=0.2,
treeAugment=TRUE,
printItrSize=100,
showProgress=TRUE,
stopError=0.01,
miniBatchSize=NA,
useBatchProgress=TRUE
){
multiLayerList<-multiLayerIntialiser(layerChoice,
unitsChoice,
networkCount,seed)
activationList<- multiActivationList(multiLayerList,activation,seed)
dataCut<-multiDataCut(x,y,networkCount,cutDataSizePercent,cutVarSizePercent,seed)
xCut=dataCut$xCut
yCut=dataCut$yCut
varCut=dataCut$varCut
multiSeed<-seed*c(1:networkCount)
modelForest<-lapply(1:networkCount, function(nc){
if(treeAugment==F){
idx<-which(sapply(1:ncol(xCut[[nc]]), function(m){
length(unique(xCut[[nc]][,m]))
})>1)
if(length(idx)>1){
xSelect<-xCut[[nc]][,idx]
}else{
xSelect<-data.frame(xCut[[nc]][,idx])
names(xSelect)<-names(xCut[[nc]])[idx]
}
deepnet(xSelect,
data.frame( yCut[[nc]]),
hiddenLayerUnits=multiLayerList[[nc]],
activationList[[nc]] ,
reluLeak,
modelType ,
iterations ,
eta ,
seed,
gradientClip,
regularisePar,
optimiser="adam",
parMomentum,
inputSizeImpact,
parRmsPropZeroAdjust,
parRmsProp,
printItrSize,
showProgress,
stopError,
miniBatchSize,
useBatchProgress,
ignoreNAerror=T)
}else{
idx<-which(sapply(1:ncol(xCut[[nc]]), function(m){
length(unique(xCut[[nc]][,m]))
})>1)
if(length(idx)>1){
xSelect<-xCut[[nc]][,idx]
}else{
xSelect<-data.frame(xCut[[nc]][,idx])
names(xSelect)<-names(xCut[[nc]])[idx]
}
deeptree( xSelect,
data.frame( yCut[[nc]]),
hiddenLayerUnits=multiLayerList[[nc]],
activationList[[nc]] ,
reluLeak,
modelType,
iterations,
eta,
seed,
gradientClip,
regularisePar,
optimiser,
parMomentum,
inputSizeImpact,
parRmsPropZeroAdjust,
parRmsProp,
treeLeaves,
treeMinSplitPercent,
treeMinSplitCount,
treeCp ,
stackPred=NA ,
printItrSize,
showProgress,
stopError,
miniBatchSize,
useBatchProgress,
ignoreNAerror=T)
} })
fitPerf<-unlist(lapply(1:length(modelForest), function(mf){
fitModel<-modelForest[[mf]]
xFit<-data.frame(xCut[[mf]][,varCut[[mf]]])
names(xFit)<-varCut[[mf]]
row.names(xFit)<-1:nrow(xFit)
yFit<-yCut[[mf]]
row.names(yFit)<-1:nrow(yFit)
if(treeAugment==T){
augmentPred<- predict.deeptree(fitModel,
newData=xFit)
if(modelType=="multiClass"){
multiClassPred<-augmentPred[,names(augmentPred)!="ypred"]
yClass<-dummy_cols(yFit)
yClass<-yClass[,names(yClass)!="y"]
names(yClass)<-str_remove_all(names(yClass),"y_")
for(idx in 1:ncol(multiClassPred)){
multiClassPred[,idx]<-as.numeric(multiClassPred[,idx])
}
selectCol<-names(multiClassPred)[names(yClass)%in%names(multiClassPred)]
sqrt(mean((
as.matrix(multiClassPred[,selectCol]-yClass[,selectCol])
)^2))/mean(as.matrix(abs(yClass)))
} else{
sqrt(mean((augmentPred$ypred-yFit$y)^2))/mean(abs(yFit$y))
}
}else{
netPred<- predict.deepnet(fitModel,
newData=xFit)
if(modelType=="multiClass"){
multiClassPred<-netPred[,names(netPred)!="ypred"]
yClass<-dummy_cols(yFit)
yClass<-yClass[,names(yClass)!="y"]
names(yClass)<-str_remove_all(names(yClass),"y_")
for(idx in 1:ncol(multiClassPred)){
multiClassPred[,idx]<-as.numeric(multiClassPred[,idx])
}
sqrt(mean((
as.matrix(multiClassPred[,names(yClass)]-yClass)
)^2))/mean(as.matrix(abs(yClass)))
}else{
sqrt(mean((netPred$ypred-yFit$y)^2))/mean(abs(yFit$y))
}
}
}))
bestFit<- min(fitPerf,na.rm = T)
errorCoverFit=bestFit+errorCover
chosenIndex<-which(!is.na(fitPerf)& fitPerf<=errorCoverFit)
chosenModels<- modelForest[chosenIndex]
deepforestmod<-list(varcut=varCut[chosenIndex],
chosenModels=chosenModels,
treeAugment=treeAugment,
modelType=modelType
)
class(deepforestmod)<-'deepforest'
return(deepforestmod)
}
Try the deepdive package in your browser
Any scripts or data that you put into this service are public.
deepdive documentation built on July 10, 2021, 5:08 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.
Defines functions deepforest
Documented in deepforest
#' @title Build or train bagged deeptree or deepnet of multiple architecture
#' @description Build or train bagged deeptree or deepnet of multiple architecture.Based on error choice either select best model or average multiple model with random variable cut,data cut and architechture
#' @param x a data frame with input variables
#' @param y a data frame with ouptut variable
#' @param networkCount Integer, Number of deepnet or deeptree to build
#' @param layerChoice vector, different layer choices
#' @param unitsChoice vector , number of units choice
#' @param cutVarSizePercent ratio, percentage of variable to for each network
#' @param cutDataSizePercent ratio, percentage of data to for each network
#' @param activation choose from "sigmoid","relu","sin","cos","none".Activations will be randomly chosen from chosen. Default is relu and sin
#' @param reluLeak numeric. Applicable when activation is "relu". Specify value between 0 any number close to zero below 1. Eg: 0.01,0.001 etc
#' @param modelType one of "regress","binary","multiClass". "regress" for regression will create a linear single unit output layer. "binary" will create a single unit sigmoid activated layer. "multiClass" will create layer with units corresponding to number of output classes with softmax activation.
#' @param iterations integer. This indicates number of iteratios or epochs in backpropagtion .The default value is 500.
#' @param eta numeric.Hyperparameter,sets the Learning rate for backpropagation. Eta determines the convergence ability and speed of convergence.
#' @param seed numeric. Set seed with this parameter. Incase of sin activation sometimes changing seed can yeild better results. Default is 2
#' @param gradientClip numeric. Hyperparameter numeric value which limits gradient size for weight update operation in backpropagation. Default is 0.8 . It can take any postive value.
#' @param regularisePar numeric. L2 Regularisation Parameter .
#' @param optimiser one of "gradientDescent","momentum","rmsProp","adam". Default value "adam"
#' @param parMomentum numeric. Applicable for optimiser "mometum" and "adam"
#' @param inputSizeImpact numeric. Adjusts the gradient size by factor of percentage of rows in input. For very small data set setting this to 0 could yeild faster result. Default is 1.
#' @param parRmsPropZeroAdjust numeric. Applicable for optimiser "rmsProp" and "adam"
#' @param parRmsProp numeric.Applicable for optimiser "rmsProp" and "adam"
#' @param treeLeaves vector.Optional , leaves numbers from externally trained tree model can be supplied here. If supplied then model will not build a explicit tree and just fit a neural network to mentioned leaves.
#' @param treeMinSplitPercent numeric. This parameter controls depth of tree setting min split count for leaf subdivision as percentage of observations. Final minimum split will be chosen as max of count calculted with treeMinSplitPercent and treeMinSplitCount. Default 0.3. Range 0 to 1.
#' @param treeMinSplitCount numeric. This parameter controls depth of tree setting min split count.Final minimum split will be chosen as max of count calculted with treeMinSplitPercent and treeMinSplitCount. Default 30
#' @param treeCp complexity parameter. \code{\link{rpart.control}}
#' @param errorCover Ratio. Deault is 0.2 i.e all models within 20 percent error of best model will be selected.
#' @param treeAugment logical. If True fits deeptree and if False fits deepnet. Default is T
#' @param printItrSize numeric. Number of iterations after which progress message should be shown. Default value 100 and for iterations below 100 atleast 5 messages will be seen
#' @param showProgress logical. True will show progress and F will not show progress
#' @param stopError Numeric. Rmse at which iterations can be stopped. Default is 0.01, can be set as NA in case all iterations needs to run.
#' @param miniBatchSize integer. Set the mini batch size for mini batch gradient
#' @param useBatchProgress logical. Applicable for miniBatch , setting T will use show rmse in Batch and F will show error on full dataset. For large dataset set T
#' @return returns model object which can be passed into \code{\link{predict.deepforest}}
#' @export
#' @import rpart
#' @import data.table
#' @importFrom stringr str_remove_all
#' @importFrom treeClust rpart.predict.leaves
#' @importFrom graphics barplot
#' @importFrom stats formula predict runif
#' @examples
#'
#' require(deepdive)
#'
#'x<-data.frame(x1=runif(10),x2=runif(10))
#'y<-data.frame(y=10*x$x1+20*x$x2+20)
#'
#'mdeepf<-deepforest(x,y,
#' networkCount=2,
#' layerChoice=c(2:3),
#' unitsChoice=c(4:10),
#' cutVarSizePercent=0.6,
#' cutDataSizePercent=0.6,
#' activation = c('relu',"sin"),
#' reluLeak=0.01,
#' modelType ='regress',
#' iterations = 10,
#' eta = 10 ^-2,
#' seed=2,
#' gradientClip=0.8,
#' regularisePar=0,
#' optimiser="adam",
#' parMomentum=0.9,
#' inputSizeImpact=1,
#' parRmsPropZeroAdjust=10^-8,
#' parRmsProp=0.9999,
#' treeLeaves=NA,
#' treeMinSplitPercent=0.3,
#' treeMinSplitCount=100,
#' treeCp=0.01 ,
#' errorCover=0.2,
#' treeAugment=TRUE,
#' printItrSize=100,
#' showProgress=TRUE,
#' stopError=0.01,
#' miniBatchSize=64,
#' useBatchProgress=TRUE)
deepforest<-function(x,y,
networkCount=3,
layerChoice=c(2:3),
unitsChoice=c(4:10),
cutVarSizePercent=0.6,
cutDataSizePercent=0.6,
activation = c('sigmoid',"sigmoid"),
reluLeak=0,
modelType ='regress',
iterations = 500,
eta = 10 ^-2,
seed=2,
gradientClip=0.8,
regularisePar=0,
optimiser="adam",
parMomentum=0.9,
inputSizeImpact=1,
parRmsPropZeroAdjust=10^-8,
parRmsProp=0.9999,
treeLeaves=NA,
treeMinSplitPercent=0.3,
treeMinSplitCount=100,
treeCp=0.01 ,
errorCover=0.2,
treeAugment=TRUE,
printItrSize=100,
showProgress=TRUE,
stopError=0.01,
miniBatchSize=NA,
useBatchProgress=TRUE
){
multiLayerList<-multiLayerIntialiser(layerChoice,
unitsChoice,
networkCount,seed)
activationList<- multiActivationList(multiLayerList,activation,seed)
dataCut<-multiDataCut(x,y,networkCount,cutDataSizePercent,cutVarSizePercent,seed)
xCut=dataCut$xCut
yCut=dataCut$yCut
varCut=dataCut$varCut
multiSeed<-seed*c(1:networkCount)
modelForest<-lapply(1:networkCount, function(nc){
if(treeAugment==F){
idx<-which(sapply(1:ncol(xCut[[nc]]), function(m){
length(unique(xCut[[nc]][,m]))
})>1)
if(length(idx)>1){
xSelect<-xCut[[nc]][,idx]
}else{
xSelect<-data.frame(xCut[[nc]][,idx])
names(xSelect)<-names(xCut[[nc]])[idx]
}
deepnet(xSelect,
data.frame( yCut[[nc]]),
hiddenLayerUnits=multiLayerList[[nc]],
activationList[[nc]] ,
reluLeak,
modelType ,
iterations ,
eta ,
seed,
gradientClip,
regularisePar,
optimiser="adam",
parMomentum,
inputSizeImpact,
parRmsPropZeroAdjust,
parRmsProp,
printItrSize,
showProgress,
stopError,
miniBatchSize,
useBatchProgress,
ignoreNAerror=T)
}else{
idx<-which(sapply(1:ncol(xCut[[nc]]), function(m){
length(unique(xCut[[nc]][,m]))
})>1)
if(length(idx)>1){
xSelect<-xCut[[nc]][,idx]
}else{
xSelect<-data.frame(xCut[[nc]][,idx])
names(xSelect)<-names(xCut[[nc]])[idx]
}
deeptree( xSelect,
data.frame( yCut[[nc]]),
hiddenLayerUnits=multiLayerList[[nc]],
activationList[[nc]] ,
reluLeak,
modelType,
iterations,
eta,
seed,
gradientClip,
regularisePar,
optimiser,
parMomentum,
inputSizeImpact,
parRmsPropZeroAdjust,
parRmsProp,
treeLeaves,
treeMinSplitPercent,
treeMinSplitCount,
treeCp ,
stackPred=NA ,
printItrSize,
showProgress,
stopError,
miniBatchSize,
useBatchProgress,
ignoreNAerror=T)
} })
fitPerf<-unlist(lapply(1:length(modelForest), function(mf){
fitModel<-modelForest[[mf]]
xFit<-data.frame(xCut[[mf]][,varCut[[mf]]])
names(xFit)<-varCut[[mf]]
row.names(xFit)<-1:nrow(xFit)
yFit<-yCut[[mf]]
row.names(yFit)<-1:nrow(yFit)
if(treeAugment==T){
augmentPred<- predict.deeptree(fitModel,
newData=xFit)
if(modelType=="multiClass"){
multiClassPred<-augmentPred[,names(augmentPred)!="ypred"]
yClass<-dummy_cols(yFit)
yClass<-yClass[,names(yClass)!="y"]
names(yClass)<-str_remove_all(names(yClass),"y_")
for(idx in 1:ncol(multiClassPred)){
multiClassPred[,idx]<-as.numeric(multiClassPred[,idx])
}
selectCol<-names(multiClassPred)[names(yClass)%in%names(multiClassPred)]
sqrt(mean((
as.matrix(multiClassPred[,selectCol]-yClass[,selectCol])
)^2))/mean(as.matrix(abs(yClass)))
} else{
sqrt(mean((augmentPred$ypred-yFit$y)^2))/mean(abs(yFit$y))
}
}else{
netPred<- predict.deepnet(fitModel,
newData=xFit)
if(modelType=="multiClass"){
multiClassPred<-netPred[,names(netPred)!="ypred"]
yClass<-dummy_cols(yFit)
yClass<-yClass[,names(yClass)!="y"]
names(yClass)<-str_remove_all(names(yClass),"y_")
for(idx in 1:ncol(multiClassPred)){
multiClassPred[,idx]<-as.numeric(multiClassPred[,idx])
}
sqrt(mean((
as.matrix(multiClassPred[,names(yClass)]-yClass)
)^2))/mean(as.matrix(abs(yClass)))
}else{
sqrt(mean((netPred$ypred-yFit$y)^2))/mean(abs(yFit$y))
}
}
}))
bestFit<- min(fitPerf,na.rm = T)
errorCoverFit=bestFit+errorCover
chosenIndex<-which(!is.na(fitPerf)& fitPerf<=errorCoverFit)
chosenModels<- modelForest[chosenIndex]
deepforestmod<-list(varcut=varCut[chosenIndex],
chosenModels=chosenModels,
treeAugment=treeAugment,
modelType=modelType
)
class(deepforestmod)<-'deepforest'
return(deepforestmod)
}
Try the deepdive package in your browser
Any scripts or data that you put into this service are public.
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.