Tyler Hunt JackStat

## animate-lm-optimization.R
library(tidyverse)
library(gganimate)
library(modelr)

options(na.action = na.warn)

slopes <- 0
intercepts <- 0
fit <- 0

## Visualizing T-Test Statistical Power.R

## Let's create a function for conducting power analysis for two groups with a t-test
t.test.power <- function(N = 100, samples = 100, g1_mean = 0, g1_sd = 1,
                         g2_mean = 0, g2_sd = 1){

  ## Setting up an empty vector to collect p-values for each sample
  PValues <- rep(NA, samples)
  Group1Mean <- rep(NA, samples)
  Group2Mean <- rep(NA, samples)
  for(i in 1:samples){

## curl-test


library(curl)


testfunc <- function(x){
  ## Setting up data vendor calls in parallel
  ###############################
  out <<- list()
  dbout <<- list()

## xgb.plot.single.tree.R
graphTree <- function(feature_names, model, tree) {

  allTrees <- xgb.model.dt.tree(feature_names = feature_names, model = model)

  allTrees <-
  allTrees %>%
    filter(Tree == tree)

  allTrees <- data.table(allTrees)


## xgboostUnwind.R
library(xgboost)
library(dplyr)
library(parallel)
library(ModelMetrics)

print(detectCores())

Train <-
  iris %>%
  mutate(

## allScoringSummary.R
#devtools::install_github("JackStat/cfb")

library(foreach)
library(cfb)

TotalWeeks = 8

allScoringSummary <- foreach(i = 1:TotalWeeks, .combine = 'rbind') %do% {
  message(i)
  scb <- NCAAscoreboard(2016, i)

## for() vs unlist(lapply())
library(microbenchmark)
SS <- runif(10000)

forLoop <- function(x){
  fill = rep(NA, length(x))
  for(i in 1:length(x)){
    fill[i] = sqrt(x[i])
  }
  return(fill)
}

## quick auc comparison
N = 100000
Actual = as.numeric(runif(N) > .5)
Predicted = as.numeric(runif(N))

actual = Actual
predicted = Predicted

s1 <- system.time(a1 <- ModelMetrics::auc(Actual, Predicted))
## Suffers from integer overflow
s2 <- system.time(a2 <- Metrics::auc(Actual, Predicted))

## vagrant_spot_instance.md

      
              1 file
            
          
              0 forks
            
          
              0 comments
            
          
              0 stars
            
          
                JackStat
                / vagrant_spot_instance.md
            
            
              Last active
              September 22, 2015 03:07
                — forked from ozzyjohnson/vagrant_spot_instance.md
            
              
                Use an EC2 spot instance with Vagrant
              
          
    Setup:
git clone git@github.com:nabeken/vagrant-aws.git
cd vagrant-aws
git checkout spot-3
gem build vagrant-aws.gemspec
vagrant install plugin vagrant-aws-0.5.0.gem

Vagrantfile:

  
## xtable_printbold.R
## 2010-06-25
## (c) Felix Andrews <felix@nfrac.org>
## GPL-2

## If 'which' is given it should be a logical matrix specifying bold cells.
## Otherwise: in each column or row with numeric data, the maximum or minimum
## value is set bold; 'max' can have entries for each column/row, NA means skip.

## Examples:
## library(xtable)
	library(tidyverse)
	library(gganimate)
	library(modelr)

	options(na.action = na.warn)

	slopes <- 0
	intercepts <- 0
	fit <- 0

	## Let's create a function for conducting power analysis for two groups with a t-test
	t.test.power <- function(N = 100, samples = 100, g1_mean = 0, g1_sd = 1,
	g2_mean = 0, g2_sd = 1){

	## Setting up an empty vector to collect p-values for each sample
	PValues <- rep(NA, samples)
	Group1Mean <- rep(NA, samples)
	Group2Mean <- rep(NA, samples)
	for(i in 1:samples){


	library(curl)


	testfunc <- function(x){
	## Setting up data vendor calls in parallel
	###############################
	out <<- list()
	dbout <<- list()
	graphTree <- function(feature_names, model, tree) {

	allTrees <- xgb.model.dt.tree(feature_names = feature_names, model = model)

	allTrees <-
	allTrees %>%
	filter(Tree == tree)

	allTrees <- data.table(allTrees)
	library(xgboost)
	library(dplyr)
	library(parallel)
	library(ModelMetrics)

	print(detectCores())

	Train <-
	iris %>%
	mutate(
	#devtools::install_github("JackStat/cfb")

	library(foreach)
	library(cfb)

	TotalWeeks = 8

	allScoringSummary <- foreach(i = 1:TotalWeeks, .combine = 'rbind') %do% {
	message(i)
	scb <- NCAAscoreboard(2016, i)
	library(microbenchmark)
	SS <- runif(10000)

	forLoop <- function(x){
	fill = rep(NA, length(x))
	for(i in 1:length(x)){
	fill[i] = sqrt(x[i])
	}
	return(fill)
	}
	N = 100000
	Actual = as.numeric(runif(N) > .5)
	Predicted = as.numeric(runif(N))

	actual = Actual
	predicted = Predicted

	s1 <- system.time(a1 <- ModelMetrics::auc(Actual, Predicted))
	## Suffers from integer overflow
	s2 <- system.time(a2 <- Metrics::auc(Actual, Predicted))
	## 2010-06-25
	## (c) Felix Andrews <felix@nfrac.org>
	## GPL-2

	## If 'which' is given it should be a logical matrix specifying bold cells.
	## Otherwise: in each column or row with numeric data, the maximum or minimum
	## value is set bold; 'max' can have entries for each column/row, NA means skip.

	## Examples:
	## library(xtable)