various CRAN check fixes

.Rbuildignore

@@ -6,3 +6,6 @@
 ^vignettes/Complex_shapes\.Rmd$
 ^vignettes/Runtime_comparison\.Rmd$
 ^\.travis\.yml$
+^Makefile$
+^CONTRIBUTORS\.md$
+^README.Rmd$

DESCRIPTION

@@ -3,12 +3,11 @@ Type: Package
 Title: Permutation Distribution Clustering
 Version: 1.0.4
 Date: 2020-02-02
-Author: Andreas M. Brandmaier [aut, cre]
 Authors@R: c(person("Andreas M. Brandmaier", 
       email="[email protected]", 
       role=c("aut","cre"),
       comment = c(ORCID = "0000-0001-8765-6982")))
-Maintainer: Andreas M. Brandmaier <[email protected].de>
+Maintainer: Andreas M. Brandmaier <andy@brandmaier.de>
 Description: Permutation Distribution Clustering is a clustering method for time series. Dissimilarity of time series is formalized as the divergence between their permutation distributions. The permutation distribution was proposed as measure of the complexity of a time series.
 License: GPL-3
 Imports: stats, 
@@ -21,4 +20,5 @@ Suggests:
     knitr,
     rmarkdown
 VignetteBuilder: knitr
-URL: http://brandmaier.github.io/pdc, https://github.com/brandmaier/pdc
+URL: https://brandmaier.github.io/pdc/, https://github.com/brandmaier/pdc
+RoxygenNote: 7.1.1

NAMESPACE

@@ -40,7 +40,7 @@ export("convert.image")
 #
 # load shared object
 #
-useDynLib(pdc)
+useDynLib(pdc,  .registration = TRUE)
 
 #
 # imports from default (but not base) packages
@@ -53,4 +53,5 @@ importFrom("graphics", "abline", "contour", "image", "lines", "par",
              "strwidth", "text")
 importFrom("stats", "as.dendrogram", "as.dist", "cmdscale", "hclust",
              "pchisq")
+importFrom("stats", "qnorm")
 importFrom("utils", "str")

R/codebook.entropy.R

@@ -1,8 +1,8 @@
 codebook.entropy <-
-function(data, m, t)
+function(data, m, t, normalize_by_observed=FALSE)
 {
 
-	ent <- entropy(codebook(data,m, t))
+	ent <- entropy(codebook(data, m, t),normalize_by_observed=normalize_by_observed)
 
 	return( ent);
 }

R/entropy.R

@@ -1,8 +1,9 @@
 entropy <-
-function(dist)
+function(dist, normalize_by_observed=FALSE)
 {
 	n <- length(dist)
 	dist <- dist[dist!=0]
+	if (normalize_by_observed) n <- length(dist)
 	if (length(dist)==1) return (0);
 	return(-sum(dist*log(dist))/log(n));
 

R/entropyHeuristic.R

@@ -6,7 +6,7 @@ function(X, m.min=3, m.max=7, t.min=1, t.max=1)
 }
 
 entropyHeuristic <-
-function(X, m.min=3, m.max=7, t.min=1, t.max=1)
+function(X, m.min=3, m.max=7, t.min=1, t.max=1, normalize_by_observed=FALSE)
 {
 	X <- as.matrix(X)
 
@@ -22,7 +22,7 @@ function(X, m.min=3, m.max=7, t.min=1, t.max=1)
 	 {
 	 	ent[k,1] <- j
 	 	ent[k,2] <- i
-		ent[k,3] <- mean(apply(FUN=codebook.entropy, MARGIN=2, X, m=i,t=j))
+		ent[k,3] <- mean(apply(FUN=codebook.entropy, MARGIN=2, X, m=i,t=j, normalize_by_observed=normalize_by_observed))
 		k <- k+1
 	 }
 	best <- which.min(ent[,3]);

R/pdclust.R

@@ -1,25 +1,23 @@
 pdclust <-
 function(X, m=NULL, t=NULL, divergence=symmetricAlphaDivergence, clustering.method="complete") 
 {
-	user.m <- !is.null(m);
-	user.t <- !is.null(t);
+	user.m <- !is.null(m)
+	user.t <- !is.null(t)
 
 	if ((is.null(m)) && (is.null(t))) {
-			m <- entropyHeuristic(X)$m;
+			m <- entropyHeuristic(X)$m
 	}
 
 	if (is.null(m)) {
-			m <- entropyHeuristic(X, t.min=t, t.max=t)$m;
+			m <- entropyHeuristic(X, t.min=t, t.max=t)$m
 	}
 
 	if (is.null(t)) {
-			t <- 1;
+			t <- 1
 	}
-
-
 
 	# calculate divergence matrix			
-	D <- pdcDist(X,m,t,divergence);
+	D <- pdcDist(X,m,t,divergence)
 
 	# start hierarchical clustering
 	if (clustering.method == "complete") {

R/udcDist.R

@@ -3,8 +3,9 @@ udConvert <- function(x)  { ifelse(sign(diff(x))==1,1,0)  }
 udCodeword <- function(x, m) {sum( 2^(1:(m-1)-1) * udConvert(x) )+1}
 
 udComplexity <- function(x, compression.type="zip") {
-  return(length(memCompress(paste0(udConvert(x),collapse = "")), 
-                type=compression.type)/length(x))
+  return(length(
+    memCompress(paste0(udConvert(x),collapse = ""),type=compression.type) 
+              )/length(x))
 }
 
 udcDist <- function(X, compression.type="zip") {

demo/00Index

@@ -0,0 +1 @@
+paired.tseries	Demonstrate clustering of pairs of time series

man/codebook.Rd

@@ -7,7 +7,7 @@ Codebook
 A codebook contains the permutation distribution of a time series.
 }
 \usage{
-codebook(x, m = 3, t = 1, use.fast=TRUE, normalized = TRUE, codeword_func = codeword)
+codebook(x, m = 3, t = 1, use.fast=TRUE, normalized = TRUE, codeword_func = NULL)
 }
 \arguments{
   \item{x}{
@@ -22,7 +22,7 @@ The embedding dimension.
 
 \item{normalized}{Normalize codebook such that it is a probability distribution.}
 
-\item{codeword_func}{Function to compute codewords.}
+\item{codeword_func}{Function to compute codewords. If NULL, the default internal function codebook is used.}
 }
 \details{
 The length of a codebook is the factorial of the embedding dimension. The elements of the

man/entropy.heuristic.Rd

@@ -16,7 +16,7 @@ Heuristic (MinE) automatically chooses an embedding dimension with an optimal
 representational entropy as proxy for representational power.
 }
 \usage{
-entropyHeuristic(X, m.min=3, m.max=7, t.min = 1, t.max = 1)
+entropyHeuristic(X, m.min=3, m.max=7, t.min = 1, t.max = 1, normalize_by_observed = FALSE)
 
 \method{print}{mine}(x, \dots)
 \method{summary}{mine}(object, \dots)
@@ -32,6 +32,8 @@ entropyHeuristic(X, m.min=3, m.max=7, t.min = 1, t.max = 1)
 \item{m.max}{Maximum embedding dimension}
 \item{t.min}{Minimum time-delay}
 \item{t.max}{Maximum time-delay}
+\item{normalize_by_observed}{Boolean. If false, entropy is normalized by dividing by the logarithm of the number of all possible patterns. If true,
+it is divided by the logarithm of the number of observed patterns.}
 \item{\dots}{Further arguments for the generic print, summary, and plot method.}
 
 \item{normalize}{Normalize values to range [0;1].}

vignettes/Getting_started.Rmd

@@ -4,7 +4,7 @@ author: "Andreas Brandmaier"
 date: "`r Sys.Date()`"
 output: rmarkdown::html_vignette
 vignette: >
-  %\VignetteIndexEntry{Vignette Title}
+  %\VignetteIndexEntry{Getting started}
   %\VignetteEngine{knitr::rmarkdown}
   %\VignetteEncoding{UTF-8}
 ---