jilmun

var1 <- ifelse(is.na(var), min(var, na.rm=T), var)
var2 <- ifelse(is.na(var), max(var, na.rm=T), var)
var3 <- ifelse(is.na(var), median(var, na.rm=T), var)
var4 <- ifelse(is.na(var), 0, var)

jilmun

var.freq <- table(var)
var.freq <- var.freq[rownames(var.freq) != ""]

# impute with mode, most frequent category
var.mode <- names(var.freq[var.freq==max(var.freq)])
var1 <- ifelse(is.na(var) | var=="", var.mode, var)

# impute with random category using sample distribution
var.filled <- var[!is.na(var) & var!=""]
var.random <- var.filled[sample(1:length(var.filled), 1)]

jilmun

wall.test <- c(2,5,1,2,3,4,7,7,6)
fn_puddle(wall.test)  # 10
wall.test <- c(2,5,1,3,1,2,1,7,7,6)
fn_puddle(wall.test)  # 17
wall.test <- c(2,5,1,3,1,2,1,7,7,0,7)
fn_puddle(wall.test)  # 24

## my initial attempt
#.. works on 1 big puddle, does not work for c(5,1,5,1,5)
fn_1puddle <- function(walls) {

jilmun

# custom tryCatch to return result and warnings -- http://stackoverflow.com/a/24569739/2271856
myTryCatch <- function(expr) {
  warn <- err <- NULL
  value <- withCallingHandlers(
    tryCatch(expr, error=function(e) {
      err <<- e
      NULL
    }), warning=function(w) {
      warn <<- w
      invokeRestart("muffleWarning")

jilmun

fn_NApattern <- function(myrow) {
  return(paste(is.na(myrow)+0, collapse=""))
}

# test function with dummy data
(df_dummy <- data.frame(X1=c(5,NA,3,NA,3,2,5,2),
                        X2=c(NA,0,3,NA,3,1,NA,2),
                        X3=c(1,5,NA,NA,6,1,NA,2),
                        X4=c(NA,NA,3,3,3,1,NA,NA)))
#   X1 X2 X3 X4

jilmun

# returns A-Z or AA-ZZ for up to 26*26=676 unique categories
anonymise <- function(mycats) {
  cnt <- length(unique(mycats))
  mycats <- factor(mycats, labels=1:cnt)
  if (cnt <= 26)
    return(LETTERS[mycats])
  newcats <- character(0)
  for (i in 1:ceiling(cnt/26))
    newcats <- c(newcats, paste0(LETTERS[i],LETTERS[1:26]))
  return(newcats[mycats])

jilmun

# create dummy data for testing
require(caret)
require(dplyr)
full <- data.frame(target = sample(c(0,1), 500, replace=T),
                   ID = 1:500,
                   v1 = sample(LETTERS[1:2], 500, replace=T),
                   v2 = sample(1:100, 500, replace=T),
                   v3 = sample(LETTERS[1:10], 500, replace=T),
                   stringsAsFactors = FALSE)
folds <- createFolds(full$target, k=5, list=TRUE, returnTrain=FALSE)

jilmun

# print everything after loop is finished
for (i in 0:101) {
  print(i)
  Sys.sleep(0.01)
}

# simplist way to print within loop
for (i in 0:101) {
  print(i)
  Sys.sleep(0.01)

jilmun

# TIC TAC TOE
# Run function ttt(n) to begin

ttt <- function (n = 3) {
  
  # set up board variables
  board <- matrix(rep(".", (n+1)^2), ncol=n+1)
  board[,1] <- c("/", 1:n)
  board[1,] <- c("/", 1:n)
  players <- c("X", "O")

jilmun

require(ggplot2) 

# prepare simpsons data ---------------------------------------------------
# code from http://suehpro.blogspot.com/2016/03/the-simpsons-as-chart.html
d1 <- data.frame(member=c(rep("Homer",3),  
                          rep("Marge",3),  
                          rep("Bart",3),  
                          rep("Lisa",2),  
                          rep("Maggie",2)),  
                 shade=c("HomerPants","HomerShirt","Skin",