Paper	: The Wild bootstrap with a small number of large clusters
Authors	: Canay, Santos, Shaikh
Journal : The Review of Economic and Statistics
Date	: November, 2019

This folder includes the R files to replicate the simulations in the paper. The are three R files:

	- Wild.Functions.R: Contains all necessary functions to simulate data and compute tests. 
	- Wild.Main.R: Main file to run simulations for Designs 1 and 2. This file generates data using `gen.data.one'. 
	- Wild.Break.R: Main file to run simulations for Designs 3 and 4. This file generates data using `gen.data.k'

The main files require inputs (seeds, design, q, n, alpha, etc). The values of this parameters used in the paper are contained in three files:

	- param-WildSize.txt: these are the values used to produce Table 1 in the paper. 
	
	- param-WildPower.txt: these are the values used to produce Table 2 in the paper. NOTE: to produce the results under the alternative hypothesis it is required to replace `beta=1' in line 11 of Wild.Main.R to `beta=0'. That is, one of the inputs of the Wild.Main.R is the true value of beta. Under the null, the true value is 1. Under the alternative, the true value is 0. 

	- param-WildBreak.txt: these are the values used to produce Table 3 in the paper. 

The outputs of the code are txt files and tex files with the rejection probabilities. They were run in the computer cluster Quest at Northwestern University during 2017. 



