Statistical Analysis for Random Objects and Non-Euclidean Data

Provides implementation of statistical methods for random objects lying in various metric spaces, which are not necessarily linear spaces. The core of this package is Fréchet regression for random objects with Euclidean predictors, which allows one to perform regression analysis for non-Euclidean responses under some mild conditions. Examples include distributions in 2-Wasserstein space, covariance matrices endowed with power metric (with Frobenius metric as a special case), Cholesky and log-Cholesky metrics, spherical data. References: Petersen, A., & Müller, H.-G. (2019) .

frechet: Statistical Analysis for Random Objects and Non-Euclidean Data

Test version of frechet package

Installation of the current development version

You can install the development version of the package in R using:

devtools::install_github("functionaldata/tFrechet")

Installation of the latest CRAN release

You can install the package in R using:

install.packages("frechet")

Load Package and Data

Once installed you can load the package with:

library(frechet)

Acknowledgment

The development of this package is supported by grants from NSF DMS-2014626, NIH ECHO and the Bill and Melinda Gates Foundation.