Optimize a vector of portfolio weights, constant in time, that results in the best terminal wealth.

1 | ```
bcrp.optim(x, maxit = 20, clean.up = TRUE, clean.up.eps = 1e-10, fast.only = FALSE)
``` |

`x` |
Time series of relative prices |

`maxit` |
Maximum number of iterations, passed to optim |

`clean.up` |
Typically some portfolio components are zero, but the optimization procedure stops with very small value. If clean.up is true, very small values are forced to zero. |

`clean.up.eps` |
If clean.up is true, the values less than clean.up.eps are forced to zero. |

`fast.only` |
Only use the fast but slightly inaccurate version of CRP optimization. The fast algorithm uses a quadratic approximation of the loss function so that quadprod can be used for an algebraic solution. |

TBD

The optimal set of portfolio weights.

Marc Delvaux

TBD, the article that introduced the quadratic approximation (Gyorfi log-optimal site)

Questions? Problems? Suggestions? Tweet to @rdrrHQ or email at ian@mutexlabs.com.

All documentation is copyright its authors; we didn't write any of that.