myfile<-"d:/Rstudio/student-mat.csv"
d1<-read.table(myfile, sep=";",header=TRUE)
library(randomForest)
## randomForest 4.6-14
## Type rfNews() to see new features/changes/bug fixes.
set.seed(123)
train <- sample(nrow(d1), nrow(d1)*0.7)
d1_train <- d1[train, ]
d1_test <- d1[-train, ]
set.seed(123)
d1_train.forest <- randomForest(G3~., data = d1_train, importance = TRUE)
d1_train.forest
##
## Call:
## randomForest(formula = G3 ~ ., data = d1_train, importance = TRUE)
## Type of random forest: regression
## Number of trees: 500
## No. of variables tried at each split: 10
##
## Mean of squared residuals: 2.965162
## % Var explained: 84.75
G3_predict <- predict(d1_train.forest, d1_train)
plot(d1_train$G3, G3_predict, main = 'Training Set',
xlab = 'G3', ylab = 'Predict')
abline(1, 1)
G3_predict <- predict(d1_train.forest, d1_test)
plot(d1_test$G3, G3_predict, main = 'Testing Set',
xlab = 'G3', ylab = 'Predict')
abline(1, 1)
summary(d1_train.forest)
## Length Class Mode
## call 4 -none- call
## type 1 -none- character
## predicted 276 -none- numeric
## mse 500 -none- numeric
## rsq 500 -none- numeric
## oob.times 276 -none- numeric
## importance 64 -none- numeric
## importanceSD 32 -none- numeric
## localImportance 0 -none- NULL
## proximity 0 -none- NULL
## ntree 1 -none- numeric
## mtry 1 -none- numeric
## forest 11 -none- list
## coefs 0 -none- NULL
## y 276 -none- numeric
## test 0 -none- NULL
## inbag 0 -none- NULL
## terms 3 terms call
importance_d1 <- d1_train.forest$importance
head(importance_d1)
## %IncMSE IncNodePurity
## school 0.0147867517 9.103365
## sex -0.0015824939 12.910086
## age 0.1133724937 69.792140
## address -0.0024275316 11.877789
## famsize -0.0221735937 11.758974
## Pstatus 0.0002671014 11.445783
importance_d1 <- data.frame(importance(d1_train.forest), check.names = FALSE)
head(importance_d1)
## %IncMSE IncNodePurity
## school 1.16989496 9.103365
## sex -0.09989218 12.910086
## age 3.18971951 69.792140
## address -0.10927656 11.877789
## famsize -1.33271245 11.758974
## Pstatus 0.02107185 11.445783
varImpPlot(d1_train.forest, n.var = min(30, nrow(d1_train.forest$importance)),
main = 'Top 30 - variable importance')
importance_d1 <- importance_d1[order(importance_d1$ IncNodePurity, decreasing = TRUE), ]
head(importance_d1)
## %IncMSE IncNodePurity
## G2 39.395021 2431.14728
## G1 22.991368 1327.37653
## absences 18.663228 318.39054
## failures 10.087547 244.64509
## Fjob -1.585402 82.32877
## age 3.189720 69.79214
write.table(importance_d1, 'importance_d1.txt', sep = '\t', col.names = NA, quote = FALSE)
set.seed(123)
d1_train.cv <- replicate(5, rfcv(d1_train[-ncol(d1_train)], d1_train$G3, cv.fold = 10, step = 1.5), simplify = FALSE)
d1_train.cv <- data.frame(sapply(d1_train.cv, '[[', 'error.cv'))
d1_train.cv$d1s <- rownames(d1_train.cv)
d1_train.cv <- reshape2::melt(d1_train.cv, id = 'd1s')
d1_train.cv$d1s <- as.numeric(as.character(d1_train.cv$d1s))
d1_train.cv.mean <- aggregate(d1_train.cv$value, by = list(d1_train.cv$d1s), FUN = mean)
head(d1_train.cv.mean, 10)
## Group.1 x
## 1 1 3.010648
## 2 2 3.445202
## 3 3 3.087011
## 4 4 2.489154
## 5 6 2.826811
## 6 9 2.935255
## 7 14 3.642242
## 8 21 4.036282
## 9 32 4.671210
library(ggplot2)
##
## 载入程辑包：'ggplot2'
## The following object is masked from 'package:randomForest':
##
## margin
ggplot(d1_train.cv.mean, aes(Group.1, x)) +
geom_line() +
theme(panel.grid = element_blank(), panel.background = element_rect(color = 'black', fill = 'transparent')) +
labs(title = '',x = 'Number of variables', y = 'Cross-validation error')
importance_d1 <- importance_d1[order(importance_d1$IncNodePurity, decreasing = TRUE), ]
importance_d1.select <- importance_d1[1:4, ]
importance_d1.select
## %IncMSE IncNodePurity
## G2 39.39502 2431.1473
## G1 22.99137 1327.3765
## absences 18.66323 318.3905
## failures 10.08755 244.6451
write.table(importance_d1.select, 'importance_d1.select.txt', sep = '\t', col.names = NA, quote = FALSE)
d1_id.select <- rownames(importance_d1.select)
d1.select <- d1[ ,c(d1_id.select, 'G3')]
d1.select <- reshape2::melt(d1.select, id = 'G3')
ggplot(d1.select, aes(x = G3, y = value)) +
geom_point() +
geom_smooth() +
facet_wrap(~variable, ncol = 3, scale = 'free_y') +
labs(title = '',x = 'G3', y = 'Relative abundance')
## `geom_smooth()` using method = 'loess' and formula 'y ~ x'
d1.select <- d1[ ,c(d1_id.select, 'G3')]
set.seed(123)
train <- sample(nrow(d1.select), nrow(d1.select)*0.7)
d1_train.select <- d1.select[train, ]
d1_test.select <- d1.select[-train, ]
set.seed(123)
d1_train.select.forest <- randomForest(G3~., data = d1_train.select, importance = TRUE)
d1_train.select.forest
##
## Call:
## randomForest(formula = G3 ~ ., data = d1_train.select, importance = TRUE)
## Type of random forest: regression
## Number of trees: 500
## No. of variables tried at each split: 1
##
## Mean of squared residuals: 3.414538
## % Var explained: 82.43
G3_predict <- predict(d1_train.select.forest, d1_train.select)
plot(d1_train.select$G3, G3_predict, main = 'Training Set',
xlab = 'G3', ylab = 'Predict')
abline(1, 1)
G3_predict <- predict(d1_train.select.forest, d1_test.select)
plot(d1_test.select$G3, G3_predict, main = 'Testing Set',
xlab = 'G3', ylab = 'Predict')
abline(1, 1)
library(plyr)
CVgroup <- function(k,datasize,seed){
cvlist <- list()
set.seed(seed)
n <- rep(1:k,ceiling(datasize/k))[1:datasize]
temp <- sample(n,datasize)
x <- 1:k
dataseq <- 1:datasize
cvlist <- lapply(x,function(x) dataseq[temp==x])
return(cvlist)
}
k <- 10
datasize <- nrow(d1)
cvlist <- CVgroup(k = k,datasize = datasize,seed = 123)
data <- d1
pred <- data.frame()
library(plyr)
library(randomForest)
m <- seq(60,500,by = 20)
for(j in m){
progress.bar <- create_progress_bar("text")
progress.bar$init(k)
for (i in 1:k){
train <- data[-cvlist[[i]],]
test <- data[cvlist[[i]],]
model <-randomForest(G3 ~.,data = d1,ntree = j)
prediction <- predict(model,subset(test,select = -G3))
randomtree <- rep(j,length(prediction))
kcross <- rep(i,length(prediction))
temp <- data.frame(cbind(subset(test,select = G3),prediction,randomtree,kcross))
pred <- rbind(pred,temp)
print(paste("Random Tree：",j))
}
}
##
|
| | 0%[1] "Random Tree： 60"
## [1] "Random Tree： 60"
## [1] "Random Tree： 60"
## [1] "Random Tree： 60"
## [1] "Random Tree： 60"
## [1] "Random Tree： 60"
## [1] "Random Tree： 60"
## [1] "Random Tree： 60"
## [1] "Random Tree： 60"
## [1] "Random Tree： 60"
##
|
| | 0%[1] "Random Tree： 80"
## [1] "Random Tree： 80"
## [1] "Random Tree： 80"
## [1] "Random Tree： 80"
## [1] "Random Tree： 80"
## [1] "Random Tree： 80"
## [1] "Random Tree： 80"
## [1] "Random Tree： 80"
## [1] "Random Tree： 80"
## [1] "Random Tree： 80"
##
|
| | 0%[1] "Random Tree： 100"
## [1] "Random Tree： 100"
## [1] "Random Tree： 100"
## [1] "Random Tree： 100"
## [1] "Random Tree： 100"
## [1] "Random Tree： 100"
## [1] "Random Tree： 100"
## [1] "Random Tree： 100"
## [1] "Random Tree： 100"
## [1] "Random Tree： 100"
##
|
| | 0%[1] "Random Tree： 120"
## [1] "Random Tree： 120"
## [1] "Random Tree： 120"
## [1] "Random Tree： 120"
## [1] "Random Tree： 120"
## [1] "Random Tree： 120"
## [1] "Random Tree： 120"
## [1] "Random Tree： 120"
## [1] "Random Tree： 120"
## [1] "Random Tree： 120"
##
|
| | 0%[1] "Random Tree： 140"
## [1] "Random Tree： 140"
## [1] "Random Tree： 140"
## [1] "Random Tree： 140"
## [1] "Random Tree： 140"
## [1] "Random Tree： 140"
## [1] "Random Tree： 140"
## [1] "Random Tree： 140"
## [1] "Random Tree： 140"
## [1] "Random Tree： 140"
##
|
| | 0%[1] "Random Tree： 160"
## [1] "Random Tree： 160"
## [1] "Random Tree： 160"
## [1] "Random Tree： 160"
## [1] "Random Tree： 160"
## [1] "Random Tree： 160"
## [1] "Random Tree： 160"
## [1] "Random Tree： 160"
## [1] "Random Tree： 160"
## [1] "Random Tree： 160"
##
|
| | 0%[1] "Random Tree： 180"
## [1] "Random Tree： 180"
## [1] "Random Tree： 180"
## [1] "Random Tree： 180"
## [1] "Random Tree： 180"
## [1] "Random Tree： 180"
## [1] "Random Tree： 180"
## [1] "Random Tree： 180"
## [1] "Random Tree： 180"
## [1] "Random Tree： 180"
##
|
| | 0%[1] "Random Tree： 200"
## [1] "Random Tree： 200"
## [1] "Random Tree： 200"
## [1] "Random Tree： 200"
## [1] "Random Tree： 200"
## [1] "Random Tree： 200"
## [1] "Random Tree： 200"
## [1] "Random Tree： 200"
## [1] "Random Tree： 200"
## [1] "Random Tree： 200"
##
|
| | 0%[1] "Random Tree： 220"
## [1] "Random Tree： 220"
## [1] "Random Tree： 220"
## [1] "Random Tree： 220"
## [1] "Random Tree： 220"
## [1] "Random Tree： 220"
## [1] "Random Tree： 220"
## [1] "Random Tree： 220"
## [1] "Random Tree： 220"
## [1] "Random Tree： 220"
##
|
| | 0%[1] "Random Tree： 240"
## [1] "Random Tree： 240"
## [1] "Random Tree： 240"
## [1] "Random Tree： 240"
## [1] "Random Tree： 240"
## [1] "Random Tree： 240"
## [1] "Random Tree： 240"
## [1] "Random Tree： 240"
## [1] "Random Tree： 240"
## [1] "Random Tree： 240"
##
|
| | 0%[1] "Random Tree： 260"
## [1] "Random Tree： 260"
## [1] "Random Tree： 260"
## [1] "Random Tree： 260"
## [1] "Random Tree： 260"
## [1] "Random Tree： 260"
## [1] "Random Tree： 260"
## [1] "Random Tree： 260"
## [1] "Random Tree： 260"
## [1] "Random Tree： 260"
##
|
| | 0%[1] "Random Tree： 280"
## [1] "Random Tree： 280"
## [1] "Random Tree： 280"
## [1] "Random Tree： 280"
## [1] "Random Tree： 280"
## [1] "Random Tree： 280"
## [1] "Random Tree： 280"
## [1] "Random Tree： 280"
## [1] "Random Tree： 280"
## [1] "Random Tree： 280"
##
|
| | 0%[1] "Random Tree： 300"
## [1] "Random Tree： 300"
## [1] "Random Tree： 300"
## [1] "Random Tree： 300"
## [1] "Random Tree： 300"
## [1] "Random Tree： 300"
## [1] "Random Tree： 300"
## [1] "Random Tree： 300"
## [1] "Random Tree： 300"
## [1] "Random Tree： 300"
##
|
| | 0%[1] "Random Tree： 320"
## [1] "Random Tree： 320"
## [1] "Random Tree： 320"
## [1] "Random Tree： 320"
## [1] "Random Tree： 320"
## [1] "Random Tree： 320"
## [1] "Random Tree： 320"
## [1] "Random Tree： 320"
## [1] "Random Tree： 320"
## [1] "Random Tree： 320"
##
|
| | 0%[1] "Random Tree： 340"
## [1] "Random Tree： 340"
## [1] "Random Tree： 340"
## [1] "Random Tree： 340"
## [1] "Random Tree： 340"
## [1] "Random Tree： 340"
## [1] "Random Tree： 340"
## [1] "Random Tree： 340"
## [1] "Random Tree： 340"
## [1] "Random Tree： 340"
##
|
| | 0%[1] "Random Tree： 360"
## [1] "Random Tree： 360"
## [1] "Random Tree： 360"
## [1] "Random Tree： 360"
## [1] "Random Tree： 360"
## [1] "Random Tree： 360"
## [1] "Random Tree： 360"
## [1] "Random Tree： 360"
## [1] "Random Tree： 360"
## [1] "Random Tree： 360"
##
|
| | 0%[1] "Random Tree： 380"
## [1] "Random Tree： 380"
## [1] "Random Tree： 380"
## [1] "Random Tree： 380"
## [1] "Random Tree： 380"
## [1] "Random Tree： 380"
## [1] "Random Tree： 380"
## [1] "Random Tree： 380"
## [1] "Random Tree： 380"
## [1] "Random Tree： 380"
##
|
| | 0%[1] "Random Tree： 400"
## [1] "Random Tree： 400"
## [1] "Random Tree： 400"
## [1] "Random Tree： 400"
## [1] "Random Tree： 400"
## [1] "Random Tree： 400"
## [1] "Random Tree： 400"
## [1] "Random Tree： 400"
## [1] "Random Tree： 400"
## [1] "Random Tree： 400"
##
|
| | 0%[1] "Random Tree： 420"
## [1] "Random Tree： 420"
## [1] "Random Tree： 420"
## [1] "Random Tree： 420"
## [1] "Random Tree： 420"
## [1] "Random Tree： 420"
## [1] "Random Tree： 420"
## [1] "Random Tree： 420"
## [1] "Random Tree： 420"
## [1] "Random Tree： 420"
##
|
| | 0%[1] "Random Tree： 440"
## [1] "Random Tree： 440"
## [1] "Random Tree： 440"
## [1] "Random Tree： 440"
## [1] "Random Tree： 440"
## [1] "Random Tree： 440"
## [1] "Random Tree： 440"
## [1] "Random Tree： 440"
## [1] "Random Tree： 440"
## [1] "Random Tree： 440"
##
|
| | 0%[1] "Random Tree： 460"
## [1] "Random Tree： 460"
## [1] "Random Tree： 460"
## [1] "Random Tree： 460"
## [1] "Random Tree： 460"
## [1] "Random Tree： 460"
## [1] "Random Tree： 460"
## [1] "Random Tree： 460"
## [1] "Random Tree： 460"
## [1] "Random Tree： 460"
##
|
| | 0%[1] "Random Tree： 480"
## [1] "Random Tree： 480"
## [1] "Random Tree： 480"
## [1] "Random Tree： 480"
## [1] "Random Tree： 480"
## [1] "Random Tree： 480"
## [1] "Random Tree： 480"
## [1] "Random Tree： 480"
## [1] "Random Tree： 480"
## [1] "Random Tree： 480"
##
|
| | 0%[1] "Random Tree： 500"
## [1] "Random Tree： 500"
## [1] "Random Tree： 500"
## [1] "Random Tree： 500"
## [1] "Random Tree： 500"
## [1] "Random Tree： 500"
## [1] "Random Tree： 500"
## [1] "Random Tree： 500"
## [1] "Random Tree： 500"
## [1] "Random Tree： 500"
maefun <- function(pred, obs) mean(abs(pred - obs))
msefun <- function(pred, obs) mean((pred - obs)^2)
nmsefun <- function(pred, obs) mean((pred - obs)^2)/mean((mean(obs) - obs)^2)
library(dplyr)
##
## 载入程辑包：'dplyr'
## The following objects are masked from 'package:plyr':
##
## arrange, count, desc, failwith, id, mutate, rename, summarise,
## summarize
## The following object is masked from 'package:randomForest':
##
## combine
## The following objects are masked from 'package:stats':
##
## filter, lag
## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, union
eval2<- pred %>% group_by(randomtree, kcross) %>%
summarise(mae = maefun(prediction, G3),
mse = msefun(prediction, G3),
nmse = nmsefun(prediction, G3),)
## `summarise()` has grouped output by 'randomtree'. You can override using the `.groups` argument.
eval2
## # A tibble: 230 x 5
## # Groups: randomtree [23]
## randomtree kcross mae mse nmse
## <dbl> <int> <dbl> <dbl> <dbl>
## 1 60 1 0.573 0.808 0.0306
## 2 60 2 0.582 0.651 0.0329
## 3 60 3 0.672 1.13 0.0448
## 4 60 4 0.547 0.639 0.0250
## 5 60 5 0.426 0.380 0.0184
## 6 60 6 0.454 0.550 0.0329
## 7 60 7 0.433 0.498 0.0294
## 8 60 8 0.661 0.982 0.0438
## 9 60 9 0.524 0.752 0.0590
## 10 60 10 0.688 0.852 0.0503
## # ... with 220 more rows
library(plyr)
CVgroup <- function(k,datasize,seed){
cvlist <- list()
set.seed(seed)
n <- rep(1:k,ceiling(datasize/k))[1:datasize]
temp <- sample(n,datasize)
x <- 1:k
dataseq <- 1:datasize
cvlist <- lapply(x,function(x) dataseq[temp==x])
return(cvlist)
}
k <- 10
datasize <- nrow(d1.select)
cvlist <- CVgroup(k = k,datasize = datasize,seed = 123)
data <- d1.select
pred <- data.frame()
library(plyr)
library(randomForest)
m <- seq(60,500,by = 20)
for(j in m){
progress.bar <- create_progress_bar("text")
progress.bar$init(k)
for (i in 1:k){
train <- data[-cvlist[[i]],]
test <- data[cvlist[[i]],]
model <-randomForest(G3~.,data=d1.select,ntree = j)
prediction <- predict(model,subset(test,select = -G3))
randomtree <- rep(j,length(prediction))
kcross <- rep(i,length(prediction))
temp <- data.frame(cbind(subset(test,select = G3),prediction,randomtree,kcross))
pred <- rbind(pred,temp)
print(paste("Random Tree：",j))
}
}
##
|
| | 0%[1] "Random Tree： 60"
## [1] "Random Tree： 60"
## [1] "Random Tree： 60"
## [1] "Random Tree： 60"
## [1] "Random Tree： 60"
## [1] "Random Tree： 60"
## [1] "Random Tree： 60"
## [1] "Random Tree： 60"
## [1] "Random Tree： 60"
## [1] "Random Tree： 60"
##
|
| | 0%[1] "Random Tree： 80"
## [1] "Random Tree： 80"
## [1] "Random Tree： 80"
## [1] "Random Tree： 80"
## [1] "Random Tree： 80"
## [1] "Random Tree： 80"
## [1] "Random Tree： 80"
## [1] "Random Tree： 80"
## [1] "Random Tree： 80"
## [1] "Random Tree： 80"
##
|
| | 0%[1] "Random Tree： 100"
## [1] "Random Tree： 100"
## [1] "Random Tree： 100"
## [1] "Random Tree： 100"
## [1] "Random Tree： 100"
## [1] "Random Tree： 100"
## [1] "Random Tree： 100"
## [1] "Random Tree： 100"
## [1] "Random Tree： 100"
## [1] "Random Tree： 100"
##
|
| | 0%[1] "Random Tree： 120"
## [1] "Random Tree： 120"
## [1] "Random Tree： 120"
## [1] "Random Tree： 120"
## [1] "Random Tree： 120"
## [1] "Random Tree： 120"
## [1] "Random Tree： 120"
## [1] "Random Tree： 120"
## [1] "Random Tree： 120"
## [1] "Random Tree： 120"
##
|
| | 0%[1] "Random Tree： 140"
## [1] "Random Tree： 140"
## [1] "Random Tree： 140"
## [1] "Random Tree： 140"
## [1] "Random Tree： 140"
## [1] "Random Tree： 140"
## [1] "Random Tree： 140"
## [1] "Random Tree： 140"
## [1] "Random Tree： 140"
## [1] "Random Tree： 140"
##
|
| | 0%[1] "Random Tree： 160"
## [1] "Random Tree： 160"
## [1] "Random Tree： 160"
## [1] "Random Tree： 160"
## [1] "Random Tree： 160"
## [1] "Random Tree： 160"
## [1] "Random Tree： 160"
## [1] "Random Tree： 160"
## [1] "Random Tree： 160"
## [1] "Random Tree： 160"
##
|
| | 0%[1] "Random Tree： 180"
## [1] "Random Tree： 180"
## [1] "Random Tree： 180"
## [1] "Random Tree： 180"
## [1] "Random Tree： 180"
## [1] "Random Tree： 180"
## [1] "Random Tree： 180"
## [1] "Random Tree： 180"
## [1] "Random Tree： 180"
## [1] "Random Tree： 180"
##
|
| | 0%[1] "Random Tree： 200"
## [1] "Random Tree： 200"
## [1] "Random Tree： 200"
## [1] "Random Tree： 200"
## [1] "Random Tree： 200"
## [1] "Random Tree： 200"
## [1] "Random Tree： 200"
## [1] "Random Tree： 200"
## [1] "Random Tree： 200"
## [1] "Random Tree： 200"
##
|
| | 0%[1] "Random Tree： 220"
## [1] "Random Tree： 220"
## [1] "Random Tree： 220"
## [1] "Random Tree： 220"
## [1] "Random Tree： 220"
## [1] "Random Tree： 220"
## [1] "Random Tree： 220"
## [1] "Random Tree： 220"
## [1] "Random Tree： 220"
## [1] "Random Tree： 220"
##
|
| | 0%[1] "Random Tree： 240"
## [1] "Random Tree： 240"
## [1] "Random Tree： 240"
## [1] "Random Tree： 240"
## [1] "Random Tree： 240"
## [1] "Random Tree： 240"
## [1] "Random Tree： 240"
## [1] "Random Tree： 240"
## [1] "Random Tree： 240"
## [1] "Random Tree： 240"
##
|
| | 0%[1] "Random Tree： 260"
## [1] "Random Tree： 260"
## [1] "Random Tree： 260"
## [1] "Random Tree： 260"
## [1] "Random Tree： 260"
## [1] "Random Tree： 260"
## [1] "Random Tree： 260"
## [1] "Random Tree： 260"
## [1] "Random Tree： 260"
## [1] "Random Tree： 260"
##
|
| | 0%[1] "Random Tree： 280"
## [1] "Random Tree： 280"
## [1] "Random Tree： 280"
## [1] "Random Tree： 280"
## [1] "Random Tree： 280"
## [1] "Random Tree： 280"
## [1] "Random Tree： 280"
## [1] "Random Tree： 280"
## [1] "Random Tree： 280"
## [1] "Random Tree： 280"
##
|
| | 0%[1] "Random Tree： 300"
## [1] "Random Tree： 300"
## [1] "Random Tree： 300"
## [1] "Random Tree： 300"
## [1] "Random Tree： 300"
## [1] "Random Tree： 300"
## [1] "Random Tree： 300"
## [1] "Random Tree： 300"
## [1] "Random Tree： 300"
## [1] "Random Tree： 300"
##
|
| | 0%[1] "Random Tree： 320"
## [1] "Random Tree： 320"
## [1] "Random Tree： 320"
## [1] "Random Tree： 320"
## [1] "Random Tree： 320"
## [1] "Random Tree： 320"
## [1] "Random Tree： 320"
## [1] "Random Tree： 320"
## [1] "Random Tree： 320"
## [1] "Random Tree： 320"
##
|
| | 0%[1] "Random Tree： 340"
## [1] "Random Tree： 340"
## [1] "Random Tree： 340"
## [1] "Random Tree： 340"
## [1] "Random Tree： 340"
## [1] "Random Tree： 340"
## [1] "Random Tree： 340"
## [1] "Random Tree： 340"
## [1] "Random Tree： 340"
## [1] "Random Tree： 340"
##
|
| | 0%[1] "Random Tree： 360"
## [1] "Random Tree： 360"
## [1] "Random Tree： 360"
## [1] "Random Tree： 360"
## [1] "Random Tree： 360"
## [1] "Random Tree： 360"
## [1] "Random Tree： 360"
## [1] "Random Tree： 360"
## [1] "Random Tree： 360"
## [1] "Random Tree： 360"
##
|
| | 0%[1] "Random Tree： 380"
## [1] "Random Tree： 380"
## [1] "Random Tree： 380"
## [1] "Random Tree： 380"
## [1] "Random Tree： 380"
## [1] "Random Tree： 380"
## [1] "Random Tree： 380"
## [1] "Random Tree： 380"
## [1] "Random Tree： 380"
## [1] "Random Tree： 380"
##
|
| | 0%[1] "Random Tree： 400"
## [1] "Random Tree： 400"
## [1] "Random Tree： 400"
## [1] "Random Tree： 400"
## [1] "Random Tree： 400"
## [1] "Random Tree： 400"
## [1] "Random Tree： 400"
## [1] "Random Tree： 400"
## [1] "Random Tree： 400"
## [1] "Random Tree： 400"
##
|
| | 0%[1] "Random Tree： 420"
## [1] "Random Tree： 420"
## [1] "Random Tree： 420"
## [1] "Random Tree： 420"
## [1] "Random Tree： 420"
## [1] "Random Tree： 420"
## [1] "Random Tree： 420"
## [1] "Random Tree： 420"
## [1] "Random Tree： 420"
## [1] "Random Tree： 420"
##
|
| | 0%[1] "Random Tree： 440"
## [1] "Random Tree： 440"
## [1] "Random Tree： 440"
## [1] "Random Tree： 440"
## [1] "Random Tree： 440"
## [1] "Random Tree： 440"
## [1] "Random Tree： 440"
## [1] "Random Tree： 440"
## [1] "Random Tree： 440"
## [1] "Random Tree： 440"
##
|
| | 0%[1] "Random Tree： 460"
## [1] "Random Tree： 460"
## [1] "Random Tree： 460"
## [1] "Random Tree： 460"
## [1] "Random Tree： 460"
## [1] "Random Tree： 460"
## [1] "Random Tree： 460"
## [1] "Random Tree： 460"
## [1] "Random Tree： 460"
## [1] "Random Tree： 460"
##
|
| | 0%[1] "Random Tree： 480"
## [1] "Random Tree： 480"
## [1] "Random Tree： 480"
## [1] "Random Tree： 480"
## [1] "Random Tree： 480"
## [1] "Random Tree： 480"
## [1] "Random Tree： 480"
## [1] "Random Tree： 480"
## [1] "Random Tree： 480"
## [1] "Random Tree： 480"
##
|
| | 0%[1] "Random Tree： 500"
## [1] "Random Tree： 500"
## [1] "Random Tree： 500"
## [1] "Random Tree： 500"
## [1] "Random Tree： 500"
## [1] "Random Tree： 500"
## [1] "Random Tree： 500"
## [1] "Random Tree： 500"
## [1] "Random Tree： 500"
## [1] "Random Tree： 500"
maefun <- function(pred, obs) mean(abs(pred - obs))
msefun <- function(pred, obs) mean((pred - obs)^2)
nmsefun <- function(pred, obs) mean((pred - obs)^2)/mean((mean(obs) - obs)^2)
library(dplyr)
eval3<- pred %>% group_by(randomtree, kcross) %>%
summarise(mae = maefun(prediction, G3),
mse = msefun(prediction, G3),
nmse = nmsefun(prediction, G3),)
## `summarise()` has grouped output by 'randomtree'. You can override using the `.groups` argument.
eval3
## # A tibble: 230 x 5
## # Groups: randomtree [23]
## randomtree kcross mae mse nmse
## <dbl> <int> <dbl> <dbl> <dbl>
## 1 60 1 1.42 4.62 0.175
## 2 60 2 1.17 2.81 0.142
## 3 60 3 1.21 3.27 0.130
## 4 60 4 1.11 2.08 0.0815
## 5 60 5 1.03 2.68 0.130
## 6 60 6 1.08 2.54 0.152
## 7 60 7 1.07 2.80 0.165
## 8 60 8 1.14 3.04 0.136
## 9 60 9 0.998 2.80 0.220
## 10 60 10 1.18 2.18 0.129
## # ... with 220 more rows