dreidpath

################################################################
# A student has five imputed datasets.  She wanted to know the #
# easiest way to repeat the analysis with each dataset and     #
# store the results.  This may not be the "easiest" way, but   #
# it is a way                                                  #
################################################################
#
################################################################
# Create five toy datasets to demonstrate the approach         #
################################################################

dreidpath

#############################################################################
# Two functions to merge a series of dataframes. This is needed if you want #
# to create a single "flat file" from multiple dataframes.                  #
#############################################################################

# Create an infix function %<<% that will merge two data frames. 
# Note Bene: 
# the dataframes to merge must have a shared column with the matched unique 
# id named "id".  You can vary this name by altering the code on line 12.
#

dreidpath

##########################################################################################################
#  Analysis of Blood glucose / HbA1C data to show that y ~ x is not the same as x ~ y.                   #
#  Author: Daniel D Reidpath, Pascale Allotey, & Mark R Diamond                                          #
#  Date: 18 October 2015                                                                                 #
##########################################################################################################
# This analysis supports a research note (doi: 10.4225/03/56239BD13C5E0)
#
#
# Load the data.  These are derived from Daramola (2012) and held on the CMU StatLib---Datasets Archive
data <- read.csv('http://lib.stat.cmu.edu/datasets/hba1c_bloodGlucose.dat', header=T, sep='\t', skip=30)

dreidpath

returnXinY <- function(lm_model, y_values){
  # This function takes a lm model expressed in terms of y ~ x, and values for y, and returns x
  # Given a model of the form y ~ beta.x + alpha, then given values for y, return x = (y-alpha)/beta
  # Crazy I know, but there you have it.
  #
  if(class(lm_model) != "lm"){
    stop("This function requires a lm mode with a single predictor variable (y ~x)")
  }
  if(length(unlist(strsplit(toString(lm_model$call[2][[1]]), " "))) > 3){
    stop("This function only works with a lm model with a single predictor variable (y ~x)")

dreidpath

require("devtools")
#  Import a funrction that will take an R lm model expressed in terms of y ~ x. Provided values for y, return "estimated" x.
source_gist("https://gist.github.com/dreidpath/6ee2bb95c6010cff3874")

x <- 0:5  # Comment out this line on the second runs, and uncomment line 6
# x <- c(1, 0, 2:5)  # Run this one second, and comment out line 5
cat("Mean of x:", mean(x))
cat("Variance of x:", var(x))

y <- 5:0

dreidpath

#####################################################################
# A series of functions to manage Malaysian National Registration   #
# Identity Card Number (NRIC) data.  The NRIC was a form            #
# YYMMDD-PB-XXXG[*], where YYMMDD are six digits representing the   #
# date of birth/registration, PB is a two digit code representing   #
# the place of birth XXX is a random digit and G is also a random   #
# digit, the if G is even the NRIC is for a female and if it is odd #
# it is for a male.                                                 #
#-------------------------------------------------------------------#
# The functions:                                                    #

dreidpath

# A Plot of age adjusted ischemic stroke annual incidence (per 100,000) in Malaysia.
# The data come from a paper by Aziz et al. (2015) published in the Journal of Stroke &
# Cerebrovascular Disease.  For a discussion see: http://wp.me/p6R8xa-Q

year <- 2010:2014
rate <- c(44.94, 49.6, 72.41, 81.94, 126.25)
model1 <- lm(rate ~ year + I(year^2))

plot(year, rate,
     xlab="Year",

dreidpath

# Data: Percentage variation within age group between the 1993 and 2016 UK populations
# The data original data from which the numbers were derived come from:
# https://www.census.gov/population/international/data/idb/region.php?N=%20Results%20&T=10&A=separate&RT=0&Y=1993,2016&R=-1&C=UK
variation <- c(-10.9589041096, -13.5802469136, -12.987012987, -8.8235294118, -4.5454545455, 
               5.7971014493, 35.1851851852, 25.4901960784, 10.2040816327, 17.0212765957, 0, 
               6.6666666667, 24.9855625305)

age_grp <- c("20-24", "25-29", "30-34", "35-39", "40-44", "45-49", "50-54", "55-59", "60-64", "65-69", "70-74", "75-79", "80+")

dreidpath

# This appeared on R-blogger, a post by "wingfeet" from 23 December 2013: http://www.r-bloggers.com/merry-christmas-2/
# I stumbled acroos it and didn't want to lose it so I placed it here
part <- list(x0=0,y0=0,x1=0,y1=1,
             branch1=NULL,branch2=NULL,extend=NULL,
             lwd=1,depth=0,col='springgreen')

par(mfrow=c(1,1),mar=c(5, 4, 4, 2) + 0.1)
segplot <- function(tree) {
  if (is.null(tree)) return()
  segments(tree$x0,tree$y0,tree$x1,tree$y1,

dreidpath

# A random-walk in flat-land.  Note that the world is a circle.

maxworld <- 1000  # Flatland is maxworld units long
p <- integer(maxworld/2)  # place p on the unit halfway between 0 and maxworld
 
dist <- c()  # dist: the distribution of "times it took to complete each iter(ation) of the randmo walk"

for(loop in 1:1000){  # Run the siumulation of random walks 1000 times
  world <- rep(0, maxworld)  # Set up the world, as maxworld-long unvisited units
  keep_going <- T  # keep_going determined whether a random walk simulation continues