jakob-r

library(paradox)
library(R6)
library(mlr3)

# Partial Least Squares Regression

LearnerRegrPls = R6Class("LearnerRegrPls",
  inherit = LearnerRegr,

  public = list(

jakob-r

library(covid19germany) # https://github.com/nevrome/covid19germany
library(data.table)
library(sf)
library(lubridate)
library(checkmate)
library(mlr3misc)
library(ggplot2)
library(plotly)
library(leaflet)

jakob-r

javascript:(function()%7Bvar%20els%20%3D%20document.getElementsByClassName('collapse')%3Bfor%20(var%20i%20%3D%201%3B%20i%20%3C%20els.length%3B%20i%2B%2B)%20%7Bels%5Bi%5D.style.setProperty('display'%2C%20'block'%2C%20'important')%3B%7D%7D)()

jakob-r

\documentclass{beamer}

\usepackage[absolute,overlay]{textpos}
\usepackage{import} 
\setbeamercolor{background canvas}{bg=gray}

\setkeys{Gin}{width=\linewidth} % default of includegraphics width

\begin{document}

jakob-r

#' @title Build ensemble of multiple learners with sampled hyperparameters.
#'
#' @description
#' Define a learner and define which hyperparameters should get sampled.
#' The ensemble will be build of multiple learners, each with different random hyperparamters.
#' The predictions for mean and se will be averaged.
#'
#' @template arg_learner
#' @param samplers (`list`)\cr
#'   A named list of functions that create the random samples:

jakob-r

Keybase proof

I hereby claim:

• I am jakob-r on github.
• I am jakobr (https://keybase.io/jakobr) on keybase.
• I have a public key ASA0uw1eePgkoFeZK9YbYrkjVOld8XKdql-qIFLAALvwWwo

To claim this, I am signing this object:

jakob-r

n = 5
x = runif(1000 * n)
mat = matrix(x, ncol = n)
mat2 = apply(mat, 1, function(x) {
  ox = order(x)
  rox = match(x, x[ox])
  dx = diff(c(0,x[ox]))
  dx[rox]
})
mat2 = t(mat2)

jakob-r

library(mlrMBO)
library(stringi)
library(jsonlite)

# read command line args (in a not very safe way)
# Script can be called like that:
# Rscript runMBO.R iters=20 time=10 seed=1
args = commandArgs(TRUE)
# defaults:
iters = 50

jakob-r

library(mlrMBO)
library(data.table)
library(ggplot2)
configureMlr(on.par.without.desc = "warn")
set.seed(1)
# Define objective function
fn = makeRosenbrockFunction(5)

# define mbo control object
ctrl = makeMBOControl(propose.points = 4, store.model.at = 0:20)

jakob-r

surrogate.learner = makeLearner("regr.randomForest", predict.type = "se")

myTrafo = function(data, target, category, na.val) {
  catf("run mytrafon on data with")
  data.notarget = data[, colnames(data) != target, drop = FALSE]
  data = cbind(dcast(data.notarget, as.formula(paste0("seq_along(", category, ")~",category)), fill = na.val, value.var = "x", drop = FALSE)[, -1, drop = FALSE], data[,colnames(data) %in% c(category, target), drop = FALSE])
  return(data)
}
wrpTrain = function(data, target, args) {
  data = myTrafo(data = data, target = target, category = args$category, na.val = args$na.val)