mcfrank

## analysis code for Fiona Lee's class project
## Psych 241, Stanford Winter 2013
library(ggplot2)
library(reshape2)
library(psych)
library(bootstrap)
library(lme4)

## for bootstrapping 95% confidence intervals
theta <- function(x,xdata,na.rm=T) {mean(xdata[x],na.rm=na.rm)}

mcfrank

# mixture distribution power sims
rm(list=ls())

# from http://stats.stackexchange.com/questions/27088/shapiro-francia-test-error
sf.testBIG=function (x) 
  
{
  DNAME <- deparse(substitute(x))
  x <- sort(x[complete.cases(x)])
  n <- length(x)

mcfrank

## this little script shows off the use of the pubmed API through bioconductor
## requires installing Biopython (using pip)
## also requires installing the DTD files for each of the Entrez API calls, 
## but the instructions for this are given when you run the script

## useful list of Entrez databases that can be queried through API
# pmc_pubmed	PubMed citations for these articles
# pmc_refs_pubmed	PubMed article citing PMC article
# pmc_pmc_cites	   PMC articles that given PMC article cites
# pmc_pmc_citedby	PMC article citing given PMC article

mcfrank

---
title: "RMarkdown for writing scientific papers: A minimal working example"
author: "Mike Frank"
date: "`r Sys.Date()`"
output: 
  html_document:
    toc: true
    number_sections: true
bibliography: mwe.bib
---

mcfrank

---
title: 'Class 6a: Blake et al. (2015) exercise'
author: "Mike Frank"
date: "February 9, 2016"
output: 
  html_document:
    toc: true
---

# Intro

mcfrank

library(purr)
library(tidyverse)
library(stringr)


# constants
locations <- 1:3
timestep <- 1
t_end <- 6
t <- seq(0, t_end, timestep)

mcfrank

mcfrank

library(tidyverse)
library(purrr)

n1 = 12
n2 = 30
nsims = 10

get_chi2 <- function(x) {
  p1 <- rbinom(1, size = n1, prob = x$p1)
  p2 <- rbinom(1, size = n2, prob = x$p2)

mcfrank

library(tidyverse)
d <- read_csv("~/Desktop/med_table_shavit.csv") %>%
  arrange(age_grp, kinship, donation, soc_dist)

# first plot
ggplot(d, 
       aes(x = soc_dist, y = med_amnt, col = age_grp)) + 
  geom_point() + 
  facet_grid(kinship ~ donation) + 
  geom_smooth(method="lm", formula = y ~ poly(x, 2), 

mcfrank

# very quick re-implementation of a loose version of Experiment 1 from 
# Twomey & Westermann 2017

# Mike Frank (mcfrank@stanford.edu)

library(autoencoder)
library(tidyverse)

# need this to evaluate pairwise distance in training sets
euclidean_pairwise <- function (x) {