jeffwong

formatDataTable = function(test) {
  colnames = names(test)
  
  # Fix numerics
  pattern = '`%s` := as.numeric( `%s` )'
  for ( field in colnames[classes == 'numeric'])
    test[ , eval( parse( text = sprintf( pattern, field,field ) ) ) ]
  
  # Fix dates
  # Can throw a warning when trying to parse NA

jeffwong

#source http://stackoverflow.com/questions/13610293/how-to-get-rjava-0-9-3-to-work-on-os-x-10-7-4-with-oracle-java-1-7

JAVA_LD_LIBRARY_PATH=/Library/Java/JavaVirtualMachines/jdk1.7.0_55.jdk/Contents/Home/jre/lib/server
JAVA_LIBS='-L/Library/Java/JavaVirtualMachines/jdk1.7.0_55.jdk/Contents/Home/jre/lib/server -ljvm'
JAVA_CPPFLAGS='-I/Library/Java/JavaVirtualMachines/jdk1.7.0_55.jdk/Contents/Home/include -I/Library/Java/JavaVirtualMachines/jdk1.7.0_55.jdk/Contents/Home/include/darwin'
R CMD javareconf

jeffwong

require(ggplot2)
require(reshape2)
m = 2
x = seq(from=0.1, to=5, by = .1)
log = log(x)
line = 1/m*(x - m) + log(m)

df = data.frame(x = x,
                log = log,
                line = line)

jeffwong

ggplot.ccf = function(x, y, lag.min=NULL, lag.max=NULL) {
  ccf.data = ccf(x,y,plot=F)
  
  indices = which(ccf.data$lag[,1,1] %in% lag.min:lag.max)
  ccf.df = data.frame(lag = ccf.data$lag[indices,1,1],
                      correlation = ccf.data$acf[indices,1,1])
  
  ggplot(ccf.df,
         aes(x = lag, y = correlation)) +
    geom_bar(stat = 'identity') +

jeffwong

fieller = function(x,y) {
  x.mean = mean(x)
  y.mean = mean(y)
  x.sd = sd(x)
  y.sd = sd(y)
  xy.cov = cov(x,y)
  
  print(sprintf("x has mean %s, y has mean %s", x.mean, y.mean))
  print(sprintf("ratio of means is %s", x.mean/y.mean))
  

jeffwong

require(ggplot2)
require(reshape2)

set.seed(100)

smallnoise = function(n) {
  rnorm(n,5,1)
}
bignoise = function(n) {
  if(rnorm(1) > 0) rnorm(n,50,10)

jeffwong

naiveTask.local = function() {
  foreach(i = 1:10000, .inorder=T) %do% {rnorm(500)}
}
naiveTask.parallel = function() {
  require(doMC)
  registerDoMC()
  foreach(i = 1:10000, .inorder=T) %dopar% {rnorm(500)}
}

batchTask.local = function() {

jeffwong

getFactorGrid = function(df) {
  factors = getFactors(df)
  df.factors = df[,factors,with=F]
  unique(df.factors)
}

getFactors = function(df) {
  classes = lapply(df, data.class)
  colnames(df)[which(classes %in% c('character', 'factor'))]
}

jeffwong

is_ec2 = function() {grepl("^[A-Za-z0-9].*-i-[A-Za-z0-9].*$", system('hostname', intern=T))}

jeffwong

nYearsAgo = function(t, n) {
  as.Date(sprintf("%s%s", as.numeric(substr(as.character(t), 1, 4)) - n, substr(as.character(t), 5, 10)), format = "%Y-%m-%d")
}