d'Aspremont, "Identifying small mean reverting portfolios"
Fu Xhiaohang, Dhanuka Ruhan, Magnin Christophe, Sinha Prashant


********************************************************************
Estimation of parameters (Covariance and A matrices)
********************************************************************

File to run: PARAMETERS.m
This file is used to get the sparse covariance and A matrices
All the other m files are functions for PARAMETERS.m

The package SeDuMi must be installed to do the covariance selection. 
Procedure to install SeDuMi
	1) Unzip the folder SeDumi
	2) run install_sedumi.m in matlab



********************************************************************
Greedy search algorithm
********************************************************************


To run the excel sheet, simply copy the parameter data [The inverse covariance matrix and the matrix A] into 
data sheet1 and data sheet2 respectively. Then go to sheet titled macro and press the calculate button. All 
the results will be reflected on the results page.



********************************************************************
Semi Definite Programing
********************************************************************

File to run: sdp.m
This file is used to calculate the mean reversion parameter, lambda, using semi definite programing
The package cvx must be installed for this code to run 
	1) unzip the folder "cvx"
	2) run cvx_setup.m

