wabarr

to-report prob-density [x mu sigma]
  ;reporter which returns the probablity density for a given value x
  ;in a normal distribution parametrized by mean = mu and standard deviation = sigma
  report (1 / (sigma * (sqrt (2 * pi))))  * (exp (-0.5 * ((x - mu) / sigma ) ^ 2))
end

wabarr

compound <- function(start, growthrate_percent, fees_percent, years=15) {
  amount <- start
  growthrate_proportion <- growthrate_percent/100
  fees_proportion <- fees_percent/100
  yearly_vals <- rep(NA, years)
  for(year in 1:years) {
    amount <- amount*(1+growthrate_proportion) - amount*fees_proportion
    yearly_vals[year] <- amount
  }
  plot(x=1:years,y=yearly_vals, pch=16, cex=2, ylab="total value in dollars", xlab="year", main=sprintf("growing at %.3f%% with fees of $%.3f%%\nstarting value: $%.2f\nending value: $%.2f",growthrate_percent, fees_percent,start, amount))

wabarr

library(factoextra)
PCA <- prcomp(iris[,1:4], scale=T)
fviz(PCA, "var")
fviz(PCA, "ind", repel=T, habillage = iris$Species, label='none')
fviz_ellipses(PCA,habillage = iris$Species, ellipse.type = "convex", geom="point") + 
  labs(title="mytitle")

wabarr

from mesa.datacollection import DataCollector

class MyDataCollector(DataCollector):
    ## subclass DataCollector to only collect data on certain agents
    ## in this case, I only report them if they are NOT alive
    ## self.alive is an attribute that I track for my agents
    def collect(self, model):
        """ Collect all the data for the given model object. """
        if self.model_reporters:
            for var, reporter in self.model_reporters.items():

wabarr

## updated with input from @richfitz and @hadley - Thanks, guys!!
export_eps <- function(filename, expression, ...){
  old <- setEPS()
  on.exit(do.call(ps.options, old), add = TRUE)

  postscript(filename, ...)
  on.exit(dev.off(), add = TRUE)

  expression
}

wabarr

kmeans testing

Andrew Barr
September 19, 2015

library(ggplot2)

groups <- factor(rep(c("A", "B"), 100))
x <- rnorm(200) 

wabarr

TimeSeries

Andrew Barr
July 6, 2015

create totally random time series

ts1 <- cumsum(rnorm(1000))
ts2 &lt;- cumsum(rnorm(1000))

wabarr

library(plotrix)
cols <- c(rainbow(6),"white")
radii <- seq(from=1, to=0.5, length.out = length(cols))
plot(0,0,ylim=c(0,9), type = "n")
lapply(1:length(cols), FUN=function(index){
  draw.circle(0,0,radius=radii[index], col=cols[index])
})

wabarr

#get the current md5sum, and use awk to strip out filename, which is reuturned as well              
#note, this assumes that md5sum_of_database_at_last_backup.txt exists and contains the hash of db at last backup
#currently no error checking

currentMD5=$(md5sum test.db | awk '{print $1}')
lastMD5=$(cat /home/user/md5sum_of_database_at_last_backup.txt)
 
if [ "$currentMD5" == "$lastMD5" ]
 
then

wabarr

library(ggplot2)
library(reshape2)

n <- 7000
theMean <- 0
theSD <- 5
A <- rbinom(n,1,c(.5,.5)) + rnorm(n,theMean,theSD)
B <- rbinom(n,1,c(.5,.5)) + rnorm(n,theMean,theSD)
C <- rbinom(n,1,c(.5,.5)) + rnorm(n,theMean,theSD)
modelID <- 1:n