markmfredrickson

library(RItools)
library(optmatch)
set.seed(20130801)
n <- 500

library(MASS)
xs <- mvrnorm(n,
              mu = c(1,2,3),
              Sigma = matrix(c(1, 0.5, 0.2,
                  0.5, 1, 0,

markmfredrickson

library(optmatch)

example(match_on)

#' @param x An ISM or blocked ISM
#' @param rs Names of rows to keep
#' @param cs Names of columns to keep
submatrix <- function(x, rs, cs) {
  subset(x,
         x@rownames %in% rs,

markmfredrickson

module Lady
where

{--
 - The Lady Tasting Tea
 - 
 - In "Statistical Methods for Research Workers" (1925) Fisher illustrates the basic
 - ideas of randomization inference in the story of a woman who claims she can
 - tell the difference between tea in which the milk is added before or after
 - the water. The experiment consists of a tray of 8 cups of tea, 4 of which

markmfredrickson

gsize <- function(n) {
  grid <- expand.grid(rep(list(1:10), n))
  object.size(grid) / 1024
}

fsize <- function(n) {
  f <- function(i) {force(i) ; function(j) { i + j}}
  flst <- lapply(1:(10^n), f)
  object.size(flst) / 1024
}

markmfredrickson

library(RItools)

# set up the potential outcomes, n = 20
Yc <- rnorm(20)
Yt <- Yc + 1 # additive effect of 2

# "randomly" sample 1/2 of the data, this works because we know how Yc is created
Z <- rep(c(T,F), 10)
R <- ifelse(Z, Yt, Yc)

markmfredrickson

(defn data->maps "Turns a vector of vectors into a seq of maps"
  [keys data]
  (map #(zipmap keys %) data))

(defn data-merge
  "Merge two data sets on the given key"
  [merge-key a b]
  (let [indexed-b (zipmap (map merge-key b) b)]
    (map #(into % (indexed-b (merge-key %))) (filter #(contains? indexed-b (merge-key %)) a))))