There are some R packages that are useful in conjunction with Portfolio Probe.

## pprobeSup

#### purpose

The most important part of `pprobeSup`

(as in Supplemental to Portfolio Probe) is the functions to perform scenario optimization.

Many of the other functions illustrate the use of random portfolios for various tasks.

There are also a few more general functions.

#### location

This is in the Portfolio Probe repository:

install.packages("pprobeSup", repos="http://www.portfolioprobe.com/R")

or alternatively:

install.packages("pprobeSup", repos="http://www.portfolioprobe.com/R", type="source")

#### dependencies

Most — but not all — of the functions depend on Portfolio Probe.

#### S+

If you use S+, then you can get it by pointing your browser at http://www.portfolioprobe.com/Splus/

Most of the functions will work in S+ but they have not been tested at all there. One or two of the utility functions won’t work in S+ as written (because of scoping issues) — but a rewrite would be easy.

## pprobeData

#### purpose

The `pprobeData`

package is a small set of almost real data that underpins many of the examples in the Portfolio Probe Cookbook.

If you have need of realistic market data that is not encumbered with legalities, you are free to use this data for whatever purpose you like.

#### location

It lives in the Portfolio Probe repository:

install.packages("pprobeData", repos="http://www.portfolioprobe.com/R")

or alternatively:

install.packages("pprobeData", repos="http://www.portfolioprobe.com/R", type="source")

#### dependencies

This is not dependent on any other package.

#### S+

If you use S+, then you can get it by pointing your browser at http://www.portfolioprobe.com/Splus/

## BurStFin

This is a package from Burns Statistics of (mostly) financial functionality.

The primary functionality is estimating variance matrices when there are more variables (assets) than observations (time points):

`var.shrink.eqcor`

— a Ledoit-Wolf shrinkage estimate towards equal correlation`factor.model.stat`

— a statistical factor model

The next most useful function is `threeDarr`

which stacks matrices into a three-dimensional array. There are some arguments to functions in Portfolio Probe that take three-dimensional arrays. For instance you can use multiple variance matrices in a problem by passing in a three-dimensional array of them.

#### location

#### dependencies

There are no dependencies with this package.

#### S+

The functions in the package are meant to be compatible with S+. You can `dump`

the functions in R and `source`

them into S+.

## BurStMisc

#### purpose

This is a collection of miscellaneous functions.

It is related to Portfolio Probe in that it contains a function for doing general numerical optimization that uses a simple genetic algorithm somewhat reminiscent of the Portfolio Probe computational engine.

But the most useful function is `corner`

which allows you to look at a little corner of a large matrix (or data frame or higher dimensional array).

#### location

#### dependencies

There are no dependencies with this package.

#### S+

The functions in the package may or may not be compatible with S+. You can `dump`

the functions in R, `source`

them into S+, and see how it goes. Note that `genopt`

is S+ compatible — that is where it was born.