johnmyleswhite

counts <- read.csv('counts.csv', header = TRUE, sep = '\t')
counts <- transform(counts, logCount = log(Count))

ggplot(counts, aes(x = Length, y = logCount)) +
  geom_point() +
  geom_smooth(method = 'lm') +
  xlab('Length of Entire Word') +
  ylab('Log Number of Occurrences of Spelling') +
  opts(title = 'How Many O\'s Does It Take to Make a GOL?')

johnmyleswhite

df <- data.frame(Fahrenheit = c(212, 32),
                 Celsius = c(100, 0))

lm.fit <- lm(Fahrenheit ~ Celsius, data = df)

summary(lm.fit)

predict(lm.fit, data.frame(Celsius = 40))

johnmyleswhite

df <- data.frame(IsSpam = c(1, 1, 0, 1),
                 MentionsViagra = c(0, 1, 0, 1),
                 MentionsNigeria = c(1, 1, 0, 0))

logistic.fit <- glm(IsSpam ~ MentionsViagra + MentionsNigeria,
                    data = df,
                    family = binomial(link = "logit"))

summary(logistic.fit)

johnmyleswhite

df <- data.frame(Fahrenheit = c(212, 102, 32),
                 Celsius = c(100, 50, 0))

lm.fit <- lm(Fahrenheit ~ Celsius, data = df)

summary(lm.fit)

predict(lm.fit, data.frame(Celsius = 40))

johnmyleswhite

# A top priority for making DataFrames useful in Julia is the development of
# good documentation and a nice API for doing plyr+reshape style operations
# in Julia. This Gist is a draft of such documentation.

load("DataFrames")
using DataFrames

load("RDatasets")

baseball = RDatasets.data("plyr", "baseball")

johnmyleswhite

    differentiate(x::Number, target::Symbol) = 0

    function differentiate(s::Symbol, target::Symbol)
        if s == target
            return 1
        else
            return 0
        end
    end

johnmyleswhite

library("ggplot2")

n.sims <- 100
max.n.vars <- 100
n.obs <- 100

res <- data.frame()

for (sim in 1:n.sims)
{

johnmyleswhite

using Distributions
using Calculus
using Benchmark

function expectation(distr::Distribution,
                   g::Function,
	                 epsilon::Real)
    f = x -> pdf(distr, x)
    endpoints = map(e -> quantile(distr, e), (epsilon, 1 - epsilon))
    integrate(x -> f(x) * g(x), endpoints[1], endpoints[2])

johnmyleswhite

##############################################################################
#
# A macro for doing delegation
#
# This macro call
# 
#     @delegate MyContainer.elems [:size, :length, :ndims, :endof]
# 
# produces this block of expressions
# 

johnmyleswhite

# Generate (x, y) data with a sparse set of active predictors
# prob controls the frequency of predictors having zero effect
function simulate_date(n::Integer, p::Integer, prob::Real)
  x = randn(n, p)
	beta = randn(p)
	for j in 1:p
		if rand() < prob
			beta[j] = 0.0
		end
	end