@@ -0,0 +1,7 @@
+# Burczynski et al. (2006) Chrohn's Disease Data Set
+# Installs the data sets from Gene Expression Omnibus (GEO) database
+source('http://bioconductor.org/biocLite.R')
+biocLite('GEOquery')
+library('GEOquery')
+geo_obj <- getGEO('GDS1615')
+
@@ -0,0 +1,9 @@
+# Burczynski et al. (2006) Chrohn's Disease Data Set
+burczynski_x <- Table(geo_obj)
+x <- unname(t(data.matrix(burczynski_x[, -c(1:2)])))
+colnames(x) <- burczynski_x[, 1]
+
+y <- Columns(geo_obj)$disease.state
+
+burczynski <- list(x = x, y = y)
+
@@ -0,0 +1,5 @@
+# Save a compressed version of the Burczynski et al. (2006) data set.
+# The 'xz' compression format will compress the data more than the
+# default 'gzip' format. However, the 'xz' takes slightly longer
+# (~2 seconds longer) than 'gzip'.
+save(burczynski, file = "burczynski.RData", compress = "xz")
@@ -31,6 +31,13 @@ source("2-clean.r")
 source("3-save.r")
 file.rename("chiaretti.RData", "../../../data/chiaretti.RData")
 
+# Burczynski (2006)
+setwd("../burczynski-2006")
+source("1-download.r")
+source("2-clean.r")
+source("3-save.r")
+file.rename("burczynski.RData", "../../../data/burczynski.RData")
+
 # Christensen (2009)
 setwd("../christensen-2009")
 source("1-download.r")