From 04842f8b78354a749e5301c28f9aa8b3b142532f Mon Sep 17 00:00:00 2001 From: Christoph Rust Date: Wed, 12 Mar 2025 08:29:05 +0100 Subject: [PATCH] doc: fix cran check notes in documentation --- DESCRIPTION | 7 +------ R/dpca.R | 7 ++++--- README.md | 2 +- man/dpca.Rd | 7 ++++--- 4 files changed, 10 insertions(+), 13 deletions(-) diff --git a/DESCRIPTION b/DESCRIPTION index 28c6f0f..3dee74b 100644 --- a/DESCRIPTION +++ b/DESCRIPTION @@ -12,12 +12,7 @@ Authors@R: c( role = c("ctb"), email = "philipp.gersing@univie.ac.at") ) -Description: Efficient implementations for estimating Generalized - Dynamic Factor Models based on the work of Mario Forni and - Marco Lippi (2001). This package currently - covers (multivariate) spectral density estimation, truncated - eigendecomposition, selection of the number of dynamic factors, - projection on common component space. +Description: Efficient implementations for estimating Generalized Dynamic Factor Models starting with the works by Forni et al. (2000) and Forni and Lippi (2001). This package currently covers (multivariate) spectral density estimation, truncated eigendecomposition, selection of the number of dynamic factors, projection on common component space. SystemRequirements: GNU make NeedsCompilation: yes Depends: R (>= 3.5.0) diff --git a/R/dpca.R b/R/dpca.R index 57ed4f8..f5e2c0a 100644 --- a/R/dpca.R +++ b/R/dpca.R @@ -1,9 +1,10 @@ #' @title Dynamic Principal Component Analysis and Dynamic Factor Model Estimation #' #' @description \code{dpca} is used to estimate a Generalized Dynamic Factor Model -#' in the spirit of Forni et al. (2000) and Forni & Lippi (2001) via dynamic principal -#' components analysis (DPCA) a la Brillinger (2001). The number of principal components -#' can be chosen in a data-driven way using the method suggested by Hallin & Liska (2007). +#' starting with the works by Forni et al. (2000) and Forni & Lippi (2001) via dynamic +#' principal components analysis (DPCA) á la Brillinger (2001). The number of principal +#' components can be chosen in a data-driven way using the method suggested by +#' Hallin & Liska (2007). #' #' \code{dpca} has a cousin \code{\link{spca}} for estimating static principal components as #' they are used in the dynamic factor model literature around Stock & Watson (2001), diff --git a/README.md b/README.md index 4519ee7..ed3a81c 100644 --- a/README.md +++ b/README.md @@ -25,7 +25,7 @@ Moreover, `dpca` implements the method to select the number of dynamic principal - [ ] Model assessment. - [ ] Port C code to modern C++. -We are aware of the R package [`freqdom`](https://cran.r-project.org/web/packages/freqdom/index.html), developed by Siegfried Hörmann and Lukas Kidzinsiki which is a pure `R` implementation. `dpca` is written mainly in `C`. Although providing a similiar interface to that of `freqdom`, `dpca` has some unique features apart from being much faster. +We are aware of the R package [`freqdom`](https://CRAN.R-project.org/package=freqdom), developed by Siegfried Hörmann and Lukas Kidzinsiki which is a pure `R` implementation. `dpca` is written mainly in `C`. Although providing a similiar interface to that of `freqdom`, `dpca` has some unique features apart from being much faster. For instance, the convoluted filter which computes the dynamic common component from the output in `freqdom` is obtained by filtering the output twice: first to get the inputs \(what in `freqdom` is called "scores" in analogy to their FDA context\) and, second these inputs are filtered again to get the dynamic common component \("KLexpansion"\). `dpca` computes the convolution in the frequency domain. The advantage of this approach is that this filter is invariant with respect to multiplications of dynamic eigenvectors by a unit-length complex number. diff --git a/man/dpca.Rd b/man/dpca.Rd index af49575..53f8885 100644 --- a/man/dpca.Rd +++ b/man/dpca.Rd @@ -77,9 +77,10 @@ also see \code{\link{select_r}}. } \description{ \code{dpca} is used to estimate a Generalized Dynamic Factor Model -in the spirit of Forni et al. (2000) and Forni & Lippi (2001) via dynamic principal -components analysis (DPCA) a la Brillinger (2001). The number of principal components -can be chosen in a data-driven way using the method suggested by Hallin & Liska (2007). +starting with the works by Forni et al. (2000) and Forni & Lippi (2001) via dynamic +principal components analysis (DPCA) á la Brillinger (2001). The number of principal +components can be chosen in a data-driven way using the method suggested by +Hallin & Liska (2007). \code{dpca} has a cousin \code{\link{spca}} for estimating static principal components as they are used in the dynamic factor model literature around Stock & Watson (2001),