In eckartbindewald/bfxapps2: R Bioinformatics Applications

knitr::opts_chunk$set(echo = TRUE)

Preliminaries

library(automl)
library(dplyr)

REGRESSION (predict Iris Sepal.Length given other parameters)

data(iris)
trainmin = 1
trainmax = 100
testmin = 101
testmax = nrow(iris) # == 150
xmat <- as.matrix(cbind(iris[,2:4], as.numeric(iris$Species)))
ymat <- iris[,1]
xmat_train <- xmat[trainmin:trainmax,]
ymat_train <- ymat[trainmin:trainmax]
xmat_test <- xmat[testmin:testmax,]
ymat_test <- ymat[testmin:testmax] # not matrix anymore but simple vector
amlmodel <- automl_train_manual(Xref = xmat_train, Yref = ymat_train,
 hpar = list(modexec = 'trainwpso', verbose = FALSE))
res <- cbind(ymat_test, automl_predict(model = amlmodel, X = xmat_test))
colnames(res) <- c('actual', 'predict')
res <- as.data.frame(res) # convert matrix to data frame
head(res)

Visualize Prediction: Correlation: r cor(res$actual,res$predict)

library(ggplot2)
res %>% ggplot(aes(x=actual, y=predict)) + geom_point() + ggpubr::theme_pubclean()

CLASSIFICATION (predict Species given other Iris parameters)

data(iris)
Xmat = iris[,1:4]
lab2pred <- levels(iris$Species)
lghlab <- length(lab2pred)  # character vector of levels to predict: "setosa"     "versicolor" "virginica" 
iris$Species <- as.numeric(iris$Species) # convert factor to integers 1,2,3 
# code to convert class ids 1,2,3 to one-hot encoding:
Ymat_help <- matrix(seq(from = 1, to = lghlab, by = 1), nrow(Xmat),
 lghlab, byrow = TRUE)
# ymat_help looks like:
#      [,1] [,2] [,3]
# [1,]    1    2    3
# [2,]    1    2    3
# [3,]    1    2    3
# [4,]    1    2    3
# [5,]    1    2    3
# [6,]    1    2    3

Ymat_help2 <- (Ymat_help == as.numeric(iris$Species))
# ymat_help2 looks like:
#      [,1]  [,2]  [,3]
# [1,] TRUE FALSE FALSE
# [2,] TRUE FALSE FALSE
# [3,] TRUE FALSE FALSE
# [4,] TRUE FALSE FALSE
# [5,] TRUE FALSE FALSE
# [6,] TRUE FALSE FALSE
# ...
# [145,] FALSE FALSE TRUE
# [146,] FALSE FALSE TRUE
# [147,] FALSE FALSE TRUE
# [148,] FALSE FALSE TRUE
# [149,] FALSE FALSE TRUE
# [150,] FALSE FALSE TRUE
Ymat <- Ymat_help2 + 0
# ymat looks like:
#     [,1] [,2] [,3]
# [1,]    1    0    0
# [2,]    1    0    0
# [3,]    1    0    0
# [4,]    1    0    0
# [5,]    1    0    0
# [6,]    1    0    0
# ...
# [145,]    0    0    1
# [146,]    0    0    1
# [147,]    0    0    1
# [148,]    0    0    1
# [149,]    0    0    1
# [150,]    0    0    1

Splitting into testing and training set

Xmat_train <- Xmat[trainmin:trainmax,]
Ymat_train <- Ymat[trainmin:trainmax,]
Xmat_test <- Xmat[testmin:testmax,]
Ymat_test <- Ymat[testmin:testmax,]

Training AutoML Models

amlmodel <- automl_train_manual(Xref = Xmat_train, Yref = Ymat_train,
                hpar = list(modexec = 'trainwpso', verbose = FALSE))

Evaluating Classification Prediction Performance

result <- cbind(Ymat_test, round(automl_predict(model = amlmodel, X = Xmat_test)))
colnames(result) <- c(paste('act',lab2pred, sep = '_'),
 paste('pred',lab2pred, sep = '_'))
head(result)

Report Accuracy

apply(result, 1, FUN=function(x) { x[1]==x[4] & x[2] == x[5] & x[3] == x[6]})

eckartbindewald/bfxapps2 documentation built on Feb. 6, 2025, 3:22 a.m.