version 1.0.4

DESCRIPTION

@@ -1,15 +1,16 @@
 Package: osc
 Type: Package
 Title: Orthodromic Spatial Clustering
-Version: 1.0.0
-Date: 2016-05-10
+Version: 1.0.4
+Date: 2019-03-22
 Depends: raster, R (>= 2.14)
-Suggests: testthat
+Suggests: testthat, maps
 Author: Steffen Kriewald, Till Fluschnik, Dominik Reusser, Diego Rybski
 Maintainer: Steffen Kriewald <[email protected]>
 Description: Allows distance based spatial clustering of georeferenced data by implementing the City Clustering Algorithm - CCA. Multiple versions allow clustering for matrix, raster and single coordinates on a plain (euclidean distance) or on a sphere (great-circle or orthodromic distance).
 License: GPL
+URL: https://www.pik-potsdam.de/~kriewald/osc/
 NeedsCompilation: yes
-Packaged: 2016-05-10 13:05:35 UTC; kriewald
+Packaged: 2019-03-22 10:34:45 UTC; kriewald
 Repository: CRAN
-Date/Publication: 2016-05-10 19:05:05
+Date/Publication: 2019-03-22 13:30:02 UTC

MD5

@@ -1,34 +1,37 @@
-f48a7147e02c0724e05b6358e127ab85 *DESCRIPTION
-33268da2b5a0174d693206835bc49e9e *NAMESPACE
+ef6b29ceb44f4c545f1c9ae0f7eb5d45 *DESCRIPTION
+6c76923a8c2910ee5d06d5f1cecd0228 *NAMESPACE
 3b7548d94355d3aac0d7a20f96f8bf3f *R/cca.R
 456b71ba9225142206fdc6d41077d4f2 *R/cca.single.R
-911fcd63d0bc1d8c0ae950319404655d *R/code_revised_cca.R
+a820b5cfb7628cc8041ade63275d950b *R/code_revised_cca.R
 ebd670ecc096df9ce8e3f53dcc0419cb *R/getPart.R
-94bdc1f82757c353cb1003e56241cdd6 *build/vignette.rds
+fc5899e593250fad044d9db5808fb1c9 *README.md
+6183df66858900ef5caf5b69906f7942 *build/vignette.rds
 6d7bc5a2d4e9a0646afa0c7e770a26bb *data/datalist
-58fd5c862c5b503e50d5245be0be11d4 *data/exampledata.rda
-e4610df58b13625c9b23bd995a55faff *data/landcover.rda
-4c396c53f51248f850075e93a4d3ddfc *data/population.rda
-971ec966bdc8fc42d3f2eab9a45fc424 *inst/doc/paper.R
-b0f652cbee75143ec695dc018de995b2 *inst/doc/paper.pdf
-bda20f004bf359c3fffe7c064e7dfc40 *inst/doc/paper.rnw
+faf8a645e7dedb1f0350588d42ace4ff *data/exampledata.rda
+4a3eeb6d9e208622e6eb7b1e715b5a0a *data/landcover.rda
+39758932230e99ea550c14d3e7bfc0b5 *data/population.rda
+1123c4ff94fe73d63a0ba479cd61f5cb *inst/doc/paper.R
+b303be4d1ef7439cced53bb5ed3f5a39 *inst/doc/paper.pdf
+8119f6039bdfd5342556eb8f51a8141d *inst/doc/paper.rnw
 71ed23ec67d3a47015db2b0b8dc06b07 *man/assign.data.Rd
-f784386850049b09d50f411c8cca098f *man/cca.Rd
+6f635bd0482bd3436984d92eaf6c8d17 *man/cca.Rd
 a48d5876310a824b38f8bedd1fe49b96 *man/coordinate.list.Rd
 caf4c194ad1c8880a0991f019062f05f *man/exampledata.Rd
 83dddfbd273fc93fc26ca39690fa42b7 *man/landcover.Rd
-0d846cb42d6a64028ee1bf6b573ce27c *man/osc.buffer.Rd
+0335607d391a966d625027a432a6a214 *man/osc.buffer.Rd
 3fefbda3fcd0894582ba971665e7c3fc *man/population.Rd
 436aff5a71c41276d06bbd6a31f2b4e6 *src/cca-core.c
-6e3121545e3e3dc8caabba7ba9200906 *src/ccaRevolution.c
+cee2d2f0410b6c393f7c93dd5486886f *src/ccaRevolution.c
+7c22b1b500cab25872924e218f1645f5 *src/registerDynamicSymbol.c
 251bdb10958e0b5ac1737af4a05d85e2 *tests/testthat.R
 35f54901471879ab9191912c70ffc422 *tests/testthat/test_cca_multi.R
 1173e8ffffc2fc75fee4a0260c93a919 *tests/testthat/test_cca_single.R
 d8266d70da0e7737622bb663ea664944 *tests/testthat/test_coordinate_list.R
 716806c191b146997b4f6aaa4bd3be08 *vignettes/compare.pdf
-bda20f004bf359c3fffe7c064e7dfc40 *vignettes/paper.rnw
-49979365c9b21512171323f8aebf9c87 *vignettes/pics/exdat1.pdf
-1fdeb0ba794228c0c0a232b901148d0c *vignettes/pics/exdat2.pdf
-8183cf69156264257790404ae1a68a15 *vignettes/pics/fig-concordance.tex
-63dcbf8e81d4a3f795ff4ce9bc7ed9be *vignettes/pics/landcover.pdf
-b273fda9b9218cf2c707d4a4a5bd8b7a *vignettes/pics/raster.pdf
+8119f6039bdfd5342556eb8f51a8141d *vignettes/paper.rnw
+d8d5c1be654100f7a714505e8431c493 *vignettes/pics/exdat1.pdf
+249bbc80805b329a9f1677bd35b8a440 *vignettes/pics/exdat2.pdf
+c8c5232df87deeb59f8f53bd479f45be *vignettes/pics/fig-concordance.tex
+4c5bf29c6181217bd3b701138ebc34f3 *vignettes/pics/landcover.pdf
+3812352fa9f9a0faf087a27284493af2 *vignettes/pics/point_data.pdf
+00657ccce1c7e2eab99949d3f9d8cbef *vignettes/pics/raster.pdf

NAMESPACE

@@ -1,4 +1,4 @@
-useDynLib(osc)
+useDynLib(osc, .registration = TRUE)
 
 export(cca, cca.single, assign.data, osc.buffer, coordinate.list)
 import(raster)

R/code_revised_cca.R

@@ -54,6 +54,9 @@ ccaRm <- function(data="", d=500, res.x=NULL, res.y=NULL, cell.class, compare)
     data <- coordinate.list(data, cell.class, compare)
   }
 
+  print(paste("use as X-coordinates column '",colnames(data)[1],"'"))
+  print(paste("use as Y-coordinates column '",colnames(data)[2],"'"))
+
   # preprocessing
   n1 <- length(data[,2]);
   data <- ccaR.order(data);
@@ -66,7 +69,14 @@ ccaRm <- function(data="", d=500, res.x=NULL, res.y=NULL, cell.class, compare)
   max_abs <- max(abs(data[,2]));
   min_dist <- acos(sin(max_abs)*sin(max_abs)+cos(max_abs)*cos(max_abs)*cos(res.y*pi/180))*6371000;
   min_dist_h <- acos(sin(0)*sin(res.x*pi/180)+cos(0)*cos(res.x*pi/180)*cos(0))*6371000;
-
+
+  # separate additional data for later
+  add.data <- NULL
+  if(ncol(data)>2){
+    add.data <- data[,3:ncol(data)]
+    data <- data[,1:2]
+  }
+
   # create place holder for cluster id
   data <- cbind(data, rep(0,length(data[,1])))
   data <- as.matrix(data)
@@ -90,6 +100,8 @@ ccaRm <- function(data="", d=500, res.x=NULL, res.y=NULL, cell.class, compare)
   mat1[,2] <- mat1[,2]*180/pi;
   mat1 <- as.data.frame(mat1)
   colnames(mat1) <- c("long","lat","cluster_id")
+  # re-add additional data
+  if(!is.null(add.data)){mat1 <- cbind(mat1, add.data)}
   return(list(cluster=mat1,size=m1$mm));
 }
 
@@ -102,12 +114,22 @@ ccaRd <- function(data="", d=500, cell.class, compare)
      data <- coordinate.list(data, cell.class, compare)
   }
 
+  print(paste("use as X-coordinates column '",colnames(data)[1],"'"))
+  print(paste("use as Y-coordinates column '",colnames(data)[2],"'"))
+
   # preprocessing
   n1 <- length(data[,2]);
   #data <- ccaR.order(data);
   data <- data[order(data[,1]),];
   print("Sorting... Done");
 
+  # separate additional data
+  add.data <- NULL
+  if(ncol(data)>2){
+    add.data <- data[,3:ncol(data)]
+    data <- data[,1:2]
+  }
+
   # create place holder for cluster id
   data <- cbind(data, rep(0,length(data[,1])))
   data <- as.matrix(data)
@@ -124,6 +146,8 @@ ccaRd <- function(data="", d=500, cell.class, compare)
 
   mat1 <- as.data.frame(mat1)
   colnames(mat1) <- c("long","lat","cluster_id")
+  # re-add additional data
+  if(!is.null(add.data)){mat1 <- cbind(mat1, add.data)}
   return(list(cluster=mat1,size=m1$mm));
 }
 
@@ -168,15 +192,43 @@ assign.data <- function(cluster, points, dist=1000){
 # buffer
 #########################################################
 
-osc.buffer <- function(input, width)
+osc.buffer <- function(input, width=max(dim(input)), return.width=F, complete=F)
 {	
   if(class(input)=="RasterLayer"){
   m <- matrix(input[], nrow=input@nrows, byrow=TRUE)
   returnraster <- T}else{
     m <- input
     returnraster <- F
   }
-	m1 <- .C("ccaBuffED", m=as.integer(m), nr=as.integer(dim(m)[1]), nc=as.integer(dim(m)[2]), sz=as.integer(width))
+	if(return.width){
+#	  print(sum(m==0))
+	  if(complete){
+	    # m1 <- .C("ccaBuffEDszS",
+	    #          m=as.integer(m),
+	    #          nr=as.integer(dim(m)[1]),
+	    #          nc=as.integer(dim(m)[2]),
+	    #          sz=as.integer(width),
+	    #          nz=as.integer(sum(m==0))) 
+	    m1 <- .C("ccaBuffEDszNN",
+	             m=as.integer(m),
+	             nr=as.integer(dim(m)[1]),
+	             nc=as.integer(dim(m)[2]),
+	             sz=as.integer(width)) 
+	  }else{
+	    m1 <- .C("ccaBuffEDszN",
+	             m=as.integer(m),
+	             nr=as.integer(dim(m)[1]),
+	             nc=as.integer(dim(m)[2]),
+	             sz=as.integer(width),
+	             nz=as.integer(sum(m==0))) 
+	  }
+	}else{
+	  m1 <- .C("ccaBuffED",
+	           m=as.integer(m),
+	           nr=as.integer(dim(m)[1]),
+	           nc=as.integer(dim(m)[2]),
+	           sz=as.integer(width))
+	}
 	m1 <- m1$m
 #	m1[which(m1<0)] <- -1
 	m1 <- matrix(m1, nrow=dim(m)[1])

README.md

@@ -0,0 +1,3 @@
+# osc: Orthodromic Spatial Clustering
+
+Allows distance based spatial clustering of georeferenced data by implementing the City Clustering Algorithm - CCA. Multiple versions allow clustering for matrix, raster and single coordinates on a plain (euclidean distance) or on a sphere (great-circle or orthodromic distance).

inst/doc/paper.R

@@ -126,7 +126,55 @@ dev.off()
 
 
 ###################################################
-### code chunk number 14: paper.rnw:205-234 (eval = FALSE)
+### code chunk number 14: paper.rnw:197-200
+###################################################
+library(maps)
+head(world.cities)
+str(world.cities)
+
+
+###################################################
+### code chunk number 15: paper.rnw:203-205
+###################################################
+city.cluster <- cca(data = world.cities[,c(5,4,1:3,6)], s = 0.5)
+
+
+
+###################################################
+### code chunk number 16: paper.rnw:209-212
+###################################################
+city.cluster.m <- cca(data = world.cities[,c(5,4,1:3,6)], s = 56*1e+03,
+                      res.x = 0.05, res.y = 0.05, unit = "m")
+
+
+
+###################################################
+### code chunk number 17: paper.rnw:215-228
+###################################################
+pdf("pics/point_data.pdf", width=8, height=4)
+par(mfrow=c(1,2))
+plot(x = city.cluster[[1]]$long,
+     y = city.cluster[[1]]$lat,
+     col=city.cluster[[1]]$cluster_id,
+     main = "d = 0.5°",
+     pch=19, cex=.1, xlab="long", ylab="lat")
+plot(x = city.cluster.m[[1]]$long,
+     y = city.cluster.m[[1]]$lat,
+     col=city.cluster.m[[1]]$cluster_id,
+     main = "d = 56 km",
+     pch=19, cex=.1, xlab="long", ylab="lat")
+dev.off()
+
+
+###################################################
+### code chunk number 18: paper.rnw:238-240
+###################################################
+  coordinates(city.cluster[[1]]) <- ~long+lat
+  proj4string(city.cluster[[1]]) <- CRS("+init=epsg:4326")
+
+
+###################################################
+### code chunk number 19: paper.rnw:252-281 (eval = FALSE)
 ###################################################
 ## library(osc)
 ## data("population")

inst/doc/paper.rnw

@@ -188,10 +188,57 @@ dev.off()
 \begin{figure}
 \centering
 \includegraphics[width=\textwidth]{pics/landcover.pdf}
-\caption{Clustering of land cover data. Left side the original data and on the right side the clutered result.}
+\caption{Clustering of land cover data. Left side the original data and on the right side the clustered result.}
 \label{fig:landcover}
 \end{figure}
 
+\subsection{Clustering point data}
+
+<<results=hide, echo=TRUE>>=
+library(maps)
+head(world.cities)
+str(world.cities)
+@
+
+<<echo=TRUE>>=
+city.cluster <- cca(data = world.cities[,c(5,4,1:3,6)], s = 0.5)
+
+@
+
+
+<<echo=TRUE>>=
+city.cluster.m <- cca(data = world.cities[,c(5,4,1:3,6)], s = 56*1e+03,
+                      res.x = 0.05, res.y = 0.05, unit = "m")
+
+@
+
+<<results=hide, echo=FALSE>>=
+pdf("pics/point_data.pdf", width=8, height=4)
+par(mfrow=c(1,2))
+plot(x = city.cluster[[1]]$long,
+     y = city.cluster[[1]]$lat,
+     col=city.cluster[[1]]$cluster_id,
+     main = "d = 0.5°",
+     pch=19, cex=.1, xlab="long", ylab="lat")
+plot(x = city.cluster.m[[1]]$long,
+     y = city.cluster.m[[1]]$lat,
+     col=city.cluster.m[[1]]$cluster_id,
+     main = "d = 56 km",
+     pch=19, cex=.1, xlab="long", ylab="lat")
+dev.off()
+@
+
+\begin{figure}
+\centering
+\includegraphics[width=\textwidth]{pics/point_data.pdf}
+\caption{CBla bla bla}
+\label{fig:landcover}
+\end{figure}
+
+<<echo=TRUE>>=
+  coordinates(city.cluster[[1]]) <- ~long+lat
+  proj4string(city.cluster[[1]]) <- CRS("+init=epsg:4326")
+@
 
 \section{Comparison of the two Versions - matrix vs. list}
 \label{sec:compare}