Type: Package
Title: Filter Covariance and Correlation Matrices with Bootstrapped-Averaged Hierarchical Ansatz
Version: 0.3.0
Date: 2020-09-21
Author: Christian Bongiorno and Damien Challet
Maintainer: Damien Challet <damien.challet@gmail.com>
Description: A method to filter correlation and covariance matrices by averaging bootstrapped filtered hierarchical clustering and boosting. See Ch. Bongiorno and D. Challet, Covariance matrix filtering with bootstrapped hierarchies (2020) <doi:10.48550/arXiv.2003.05807> and Ch. Bongiorno and D. Challet, Reactive Global Minimum Variance Portfolios with k-BAHC covariance cleaning (2020) <doi:10.48550/arXiv.2005.08703>.
License: GPL-2 | GPL-3 [expanded from: GPL]
Depends: R (≥ 3.5.0), fastcluster, matrixStats
Encoding: UTF-8
LazyData: true
RoxygenNote: 7.1.0
NeedsCompilation: no
Packaged: 2020-09-21 15:57:33 UTC; damien
Repository: CRAN
Date/Publication: 2020-09-21 16:40:02 UTC

Compute the BAHC correlation matrix.

Description

Compute the BAHC correlation matrix.

Usage

filterCorrelation(x, k = 1, Nboot = 100)

Arguments

x

A matrix: x_{i,f} is feature f of object i

k

The order of filtering. k=1 corresponds to BAHC.

Nboot

The number of bootstrap copies

Value

The BAHC-filtered correlation matrix of x.

Examples

r=matrix(rnorm(1000),nrow=20)   # 20 objects, 50 features each
Cor_bahc=filterCorrelation(r)

Compute the BAHC covariance matrix.

Description

Compute the BAHC covariance matrix.

Usage

filterCovariance(x, k = 1, Nboot = 100)

Arguments

x

A matrix: x_{i,f} is feature f of object i

k

The order of filtering. k=1 corresponds to BAHC.

Nboot

The number of bootstrap copies

Value

The BAHC-filtered correlation matrix of x.

Examples

r=matrix(rnorm(1000),nrow=20)   # 20 objects, 50 features each
sigma=exp(runif(20))
rs=t(sigma %*% r) %*% sigma
Cov_bahc=filterCovariance(rs)

mirror server hosted at Truenetwork, Russian Federation.