Burns Statistics -- Public Domain Code


	Public Domain Code The sections of this page are: A Note on R and S-PLUS Financial Code General Utility Code A Note on R and S-PLUS R is free, open-source software that provides an environment for data analysis, graphics and statistics. You can download the software, documentation and so on from the R Project website. Almost everyone will want to install a binary version (as opposed to building it from source code). S-PLUS is commercial software. Both R and S-PLUS are implementations of the S language, and a great deal of code will run under both. A brief (8 page) introduction to using R and S-PLUS is "A Guide for the Unwilling S User". An exposition of some of the merits of the S language is "An Introduction to the S Language". Financial Code To use the R functions in this section: 1) Install R as described above (it's actually very easy). 2) Save the files ending in ".R" to some location on your computer. 3) Start R. 4) Source the ".R" files into R. For example, at the R prompt type: source("alpha.proxy.R") (followed by hitting the enter key). If you are on Windows, you may find it convenient to use the "Change dir" item in the File menu. 5) Read the help file. It will probably be easiest just to read help files in html -- either here or copying the html files to your own computer. There is an R package called "evalstrat" that contains a few functions for backtesting trading strategies. These were used for the computations in the Working Paper "Random Portfolios for Evaluating Trading Strategies". These functions will be of interest mostly as examples. To get the package, within R you can do: install.packages('evalstrat', repos='http://www.burns-stat.com/pages/Freecode') library(evalstrat) Alternatively, you can get one of the files containing the package: http://www.burns-stat.com/pages/Freecode/src/ contrib/evalstrat_1.0.tar.gz http://www.burns-stat.com/pages/Freecode/bin/ windows/contrib/2.2/evalstrat_1.0.zip "factor.model.stat" is an R and S-PLUS function that computes a variance matrix based on a factor model. This can estimate a variance model even when there are more variables than observations. R (and S-PLUS) code: factor.model.stat.q R help file: factor.model.stat.Rd R (and S-PLUS) code: fitted.statfacmodBurSt.q R help file: fitted.statfacmodBurSt.Rd "alpha.proxy" is an R function that computes (and possibly plots) the alpha proxy as described in the Working Paper "Portfolio Sharpening". The computations in this function will work in S-PLUS as well, but the plotting will need modification. R code: alpha.proxy.R R help file: alpha.proxy.Rd Help file in html: alpha.proxy.html "partial.rainbow" is an R function that is used to describe the colors for a contour plot. It was created for use with "alpha.proxy" but is of more general use. R code: partial.rainbow.R R help file: partial.rainbow.Rd Help file in html: partial.rainbow.html There is also some R code associated with the Technical Analysis Challenge. Here is the access to the code. General Utility Code The following pieces of code along with their help files are placed in the public domain by Burns Statistics. The functions are designed to run in both S-PLUS and R. Functions for random permutation tests: S code: permtest.R R help file: permutation.test.discrete.Rd html help file: permutation.test.discrete.html R help file: plot.permtstBurSt.Rd html help file: plot.permtstBurSt.html "Permuting Super Bowl Theory" is a Working Paper which uses this code. R for the Super Bowl is instructive for using the functions. "genopt" is a genetic algorithm for optimization that is discussed in S Poetry. genopt.R "corner" returns the selected corner of a matrix or multidimensional array. Though occasionally useful inside a function, it is most often useful for viewing objects in interactive use. (This presumes that "head" and "tail" exist if it is called using non-arrays.) S code: corner.S (updated 2002 December 17) R help file: corner.Rd Help file in html: corner.html "head" returns the first few items of an object -- the first elements of a vector, the first rows of a matrix or data frame, the first lines of a function. S code: head.S R help file: head.Rd Help file in html: head.html "tail" returns the last few items of an object -- the last elements of a vector, the last rows of a matrix or data frame, the last lines of a function. S code: tail.S R help file: tail.Rd Help file in html: tail.html Go to Burns Statistics Home. Last Modified: 2008 March 01 Direct access to this page is http://www.burns-stat.com/pages/public.html