Fixed lots of code.

DESCRIPTION

@@ -1,6 +1,6 @@
 Package: stream
-Version: 1.3-0.1
-Date: 2018-xx-xx
+Version: 1.3-1
+Date: 2019-06-07
 Title: Infrastructure for Data Stream Mining
 Authors@R: c(person("Michael", "Hahsler", role = c("aut", "cre", "cph"),
                 email = "mhahsler@lyle.smu.edu"),
@@ -17,3 +17,4 @@ URL: https://github.com/mhahsler/stream
 BugReports: https://github.com/mhahsler/stream/issues
 LinkingTo: Rcpp, BH
 License: GPL-3
+RoxygenNote: 6.1.1

NEWS.md

@@ -1,26 +1,34 @@
-# stream 1.3-0.1 (xx/xx/18)
+# stream 1.3-1 (06/07/19)
 
+## New Features
 * Added DSC_evoStream and DSC_EA. Code by Matthias Carnein.
+
+## Changes
 * Package animation is now only suggested since it requires package magick 
     which may need the imagemagick++ libraries installed. 
 
 # stream 1.3-0 (05/31/18)
 
+## New Features
 * Added DSC_BIRCH. Code and Interface by Dennis Assenmacher and Matthias Carnein.
 * Added DSC_BICO. Code by Hendrik Fichtenberger, Marc Gille, Melanie Schmidt, 
     Chris Schwiegelshohn, Christian Sohler and Interface provided by Matthias 
     Carnein and Dennis Assenmacher.
+* animate_cluster: noise now accepts "class" or "exclude" ("ignore" is deprecated).
+
+## Bug Fixes
 * DSD_ReadCSV: Fixed bug with streams that have no class/cluster label 
     (reported by Matthias Carnein).
-* animate_cluster: noise now accepts "class" or "exclude" ("ignore" is deprecated).
 
 # stream 1.2-4 (02/25/17)
 
+## Bug Fixes
 * Use dbFetch in DSD_ReadDB (new version of RSQLite).
 * Register native C routines.
 
 # stream 1.2-3 (08/07/16)
 
+## Bug Fixes
 * fixed saveDSC for DBStream.
 * fixed handling of data with d=1 (reported by Ilana Lichtenstein).
 * plot now automatically determines if the data supports a class attribute.

R/DSC_BICO.R

@@ -17,23 +17,6 @@
 # 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
 
 
-#' BICO - Fast computation of k-means coresets in a data stream
-#'
-#' BICO maintains a tree which is inspired by the clustering tree of BIRCH,
-#' a SIGMOD Test of Time award-winning clustering algorithm.
-#' Each node in the tree represents a subset of these points. Instead of
-#' storing all points as individual objects, only the number of points,
-#' the sum and the squared sum of the subset's points are stored as key features
-#' of each subset. Points are inserted into exactly one node.
-#'
-#' In this implementation, the nearest neighbour search on the first level
-#' of the tree ist sped up by projecting all points to random 1-d subspaces.
-#' The first estimation of the optimal clustering cost is computed in a
-#' buffer phase at the beginning of the algorithm.
-#'
-#' This implementation interfaces the original C++ implementation available here: \url{http://ls2-www.cs.tu-dortmund.de/grav/de/bico}.
-#' For micro-clustering, the algorithm computes the coreset of the stream. Reclustering is performed by using the \code{kmeans++} algorithm on the coreset.
-
 DSC_BICO <- function(k=5, space=10, p=10, iterations=10) {
 
   BICO <- BICO_R$new(k, space, p, iterations)

R/DSC_EA.R

@@ -53,7 +53,8 @@
 #' reset_stream(stream)
 #' plot(two, stream, type="both")
 #'
-#' ## if we have time, evaluate additional generations. This can be called at any time, also between observations.
+#' ## if we have time, evaluate additional generations. This can be
+#' ## called at any time, also between observations.
 #' two$macro_dsc$RObj$recluster(2000)
 #'
 #' ## plot improved result
@@ -85,30 +86,14 @@ DSC_EA <- function(k, generations=2000, crossoverRate=.8, mutationRate=.001, pop
 }
 
 
-
-#' Reference Class EA_R
-#'
-#' Reference class used for Reclustering using an evolutionary algorithm
-#'
-#' @field crossoverRate cross-over rate for the evolutionary algorithm
-#' @field mutationRate mutation rate for the evolutionary algorithm
-#' @field populationSize number of solutions that the evolutionary algorithm maintains
-#' @field k number of macro-clusters
-#' @field generations number of EA generations performed during reclustering
-#' @field data micro-clusters to recluster
-#' @field weights weights of the micro-clusters
-#' @field C exposed C class
-#'
-#' @author Matthias Carnein \email{matthias.carnein@@uni-muenster.de}
-#'
 EA_R <- setRefClass("EA",
                     fields = list(
-                      crossoverRate	        = "numeric",
-                      mutationRate	    = "numeric",
+                      crossoverRate	= "numeric",
+                      mutationRate	= "numeric",
                       populationSize    = "integer",
-                      k  = "integer",
+                      k           = "integer",
                       data        = "data.frame",
-                      weights	    = "numeric",
+                      weights	  = "numeric",
                       generations = "integer",
                       C = "ANY"
                     ),

R/DSC_evoStream.R

@@ -45,13 +45,13 @@
 #' @references Carnein M. and Trautmann H. (2018), "evoStream - Evolutionary Stream Clustering Utilizing Idle Times", Big Data Research.
 #'
 #' @examples
-#' stream <- DSD_Memory(DSD_Gaussians(k = 3, d = 2), 1000)
+#' stream <- DSD_Memory(DSD_Gaussians(k = 3, d = 2), 500)
 #'
 #' ## init evoStream
-#' evoStream <- DSC_evoStream(r=0.05, k=3, incrementalGenerations=1, reclusterGenerations=1000)
+#' evoStream <- DSC_evoStream(r=0.05, k=3, incrementalGenerations=1, reclusterGenerations=500)
 #'
 #' ## insert observations
-#' update(evoStream, stream, n = 1000)
+#' update(evoStream, stream, n = 500)
 #'
 #' ## micro clusters
 #' get_centers(evoStream, type="micro")
@@ -69,9 +69,11 @@
 #' reset_stream(stream)
 #' plot(evoStream, stream, type = "both")
 #'
-#' ## if we have time, evaluate additional generations. This can be called at any time, also between observations.
-#' ## by default, 1 generation is evaluated after each observation and 1000 generations during reclustering (parameters)
-#' evoStream$RObj$recluster(2000)
+#' ## if we have time, evaluate additional generations.
+#' ## This can be called at any time, also between observations.
+#' ## by default, 1 generation is evaluated after each observation and
+#' ## 1000 generations during reclustering but we set it here to 500
+#' evoStream$RObj$recluster(500)
 #'
 #' ## plot improved result
 #' reset_stream(stream)
@@ -94,19 +96,6 @@ DSC_evoStream <- function(r, lambda=0.001, tgap=100, k=2, crossoverRate=.8, muta
 }
 
 
-
-
-
-
-
-#' Reference Class evoStream_R
-#'
-#' Reference class mostly used to expose the C class object
-#'
-#' @field C exposed C class
-#'
-#' @author Matthias Carnein \email{matthias.carnein@@uni-muenster.de}
-#'
 evoStream_R <- setRefClass("evoStream_R", fields = list(
   C ="ANY"
 ))

inst/CITATION

@@ -17,13 +17,6 @@ bibentry(bibtype = "Article",
   volume       = "76",
   number       = "14",
   pages        = "1--50",
-  doi          = "10.18637/jss.v076.i14",
-
-  header       = "To cite stream in publications use:",
-  textVersion  =
-  paste("Michael Hahsler, Matthew Bolanos, John Forrest (2017).",
-        "Introduction to stream: An Extensible Framework for Data Stream Clustering Research with R.",
-        "Journal of Statistical Software, 76(14), 1-50.",
-        "doi:10.18637/jss.v076.i14")
-)
+  doi          = "10.18637/jss.v076.i14"
+  )
 

man/DSC_BICO.Rd

@@ -28,16 +28,13 @@ In this implementation, the nearest neighbour search on the first level
 of the tree ist sped up by projecting all points to random 1-d subspaces.
 The first estimation of the optimal clustering cost is computed in a
 buffer phase at the beginning of the algorithm.
-
 This implementation interfaces the original C++ implementation available here: \url{http://ls2-www.cs.tu-dortmund.de/grav/de/bico}.
 For micro-clustering, the algorithm computes the coreset of the stream. Reclustering is performed by using the \code{kmeans++} algorithm on the coreset.
 }
 \examples{
 stream <- DSD_Gaussians(k = 3, d = 2)
-
 BICO <- DSC_BICO(k = 3, p = 10, space = 100, iterations = 10)
 update(BICO, stream, n = 500)
-
 plot(BICO,stream, type = "both")
 }
 \references{
@@ -47,7 +44,6 @@ Hendrik Fichtenberger, Marc Gille, Melanie Schmidt, Chris Schwiegelshohn, Christ
 R-Interface:
 Matthias Carnein (\email{Matthias.Carnein@uni-muenster.de}),
 Dennis Assenmacher.
-
 C-Implementation:
 Hendrik Fichtenberger,
 Marc Gille,

man/DSC_TwoStage.Rd

@@ -5,21 +5,21 @@
 single process.}
 
 \usage{
-DSC_TwoStage(micro, macro) 
+DSC_TwoStage(micro, macro)
 }
 
 \arguments{
-	\item{micro}{Clustering algorithm for online stage (\code{DSC_micro})} 
-	\item{macro}{Clustering algorithm for offline stage (\code{DSC_macro})} 
+	\item{micro}{Clustering algorithm used in the online stage (\code{DSC_micro})}
+	\item{macro}{Clustering algorithm used for reclustering in the offline stage (\code{DSC_macro})}
 }
 
 \details{
-\code{update()} runs the micro-clustering stage and if centers/weights are 
-requested the reclustering is automatically performed.
+\code{update()} runs the micro-clustering stage and only when macro cluster
+centers/weights are requested, then the offline stage reclustering is automatically performed.
 }
 
 \value{
-An object of class \code{DSC_TwoStage} (subclass of  \code{DSC}, \code{DSC_Macro}). 
+An object of class \code{DSC_TwoStage} (subclass of  \code{DSC}, \code{DSC_Macro}).
 }
 
 %\references{ }
@@ -39,13 +39,13 @@ stream <- DSD_Gaussians(k=3)
 # Create a clustering process that uses a window for the online stage and
 # k-means for the offline stage (reclustering)
 win_km <- DSC_TwoStage(
-  micro=DSC_Window(horizon=100), 
+  micro=DSC_Window(horizon=100),
   macro=DSC_Kmeans(k=3)
-  ) 
+  )
 win_km
-  
-update(win_km, stream, 200) 
+
+update(win_km, stream, 200)
 win_km
-plot(win_km, stream, type="both")  
+plot(win_km, stream, type="both")
 evaluate(win_km, stream, assign="macro")
 }

src/BICO.cpp

@@ -1,30 +1,30 @@
 #include <Rcpp.h>
 using namespace Rcpp;
 
-#include <iostream>
-#include <sstream>
-#include <fstream>
-#include <random>
-#include <ctime>
-#include <time.h>
+//#include <iostream>
+//#include <sstream>
+//#include <fstream>
+//#include <random>
+//#include <ctime>
+//#include <time.h>
 
-#include <boost/algorithm/string.hpp>
+//#include <boost/algorithm/string.hpp>
 
 #include "BICO/l2metric.h"
 #include "BICO/squaredl2metric.h"
 #include "BICO/point.h"
 #include "BICO/pointweightmodifier.h"
 #include "BICO/bico.h"
-#include "BICO/randomness.h"
-#include "BICO/randomgenerator.h"
+//#include "BICO/randomness.h"
+//#include "BICO/randomgenerator.h"
 #include "BICO/proxysolution.h"
 #include "BICO/pointcentroid.h"
 #include "BICO/pointweightmodifier.h"
 #include "BICO/realspaceprovider.h"
 
 #include "BICO/master.h"
-#include <stdio.h>
-#include <stdlib.h>
+//#include <stdio.h>
+//#include <stdlib.h>
 
 
 // data structure adapted to Rcpp NumericMatrix and IntegerVector
@@ -115,21 +115,24 @@ class BICO {
   }
 
 
-  void cluster(Rcpp::NumericMatrix data){
+  void cluster(Rcpp::NumericMatrix data) {
 
     // initialize
     if(bico==NULL){
       this->d = data.ncol();
       // parameter n is not used, therefore pass dummy 0 value
       this->bico = new CluE::Bico<CluE::Point>(this->d, 0, this->k, this->p, this->space, &this->metric, &this->modifier);
+    }else{
+      if(d != data.ncol()) {
+        Rf_error("Dimensions of new data do not match the current BICO clustering.");
+      }
     }
 
     this->upToDate=false;
 
-    time_t starttime;
-
+//    time_t starttime;
     // Randomness::initialize(seed);
-    time(&starttime);
+//    time(&starttime);
     int n = data.nrow();
 
     for(int row=0; row < n; row++){
@@ -162,12 +165,12 @@ class BICO {
 
     this->micro = micro;
     this->microWeight = microWeight;
-  }
+}
 
 
 
 
-  void recluster(){
+  void recluster() {
 
     if(this->micro.nrow() == 0){
       return;
@@ -206,7 +209,7 @@ class BICO {
     this->macro = macro;
     this->macroWeight = macroWeight;
     this->assignment = assignment;
-  }
+}
 
 
   Rcpp::NumericMatrix get_microclusters(){