Add VT2-examples: all 6 chapters of 'Visualize This' (110 files) — data scraping, ETL, trends, categorical, multivariate, spatial, distribution analysis

VT2-examples/vt2-ch8-codedata-dir/make-distributions.R

@@ -0,0 +1,84 @@
+library(readr)
+library(dplyr)
+
+
+#
+# Load data
+#
+
+players <- read_csv("data/nba-players-2013-14.csv")
+head(players)
+
+
+#
+# Player heights distributions
+#
+
+warriors <- players %>%
+    filter(Team == "Warriors") %>%
+    arrange(-Ht_inches)
+
+# Bar chart
+par(mar=c(5,10,5,5))
+barplot(warriors$Ht_inches, 
+        names.arg=warriors$Name, 
+        horiz=TRUE, 
+        border=NA, las=1, 
+        main="Heights of Golden State Warriors, 2013-14")
+
+
+# Dots on a continuous height scale
+htrange <- range(warriors$Ht_inches)
+cnts <- rep(0, htrange[2]-htrange[1]+1)
+y <- c()
+for (i in 1:dim(warriors)[1]) {
+    cntindex <- warriors$Ht_inches[i] - htrange[1] + 1
+    cnts[cntindex] <- cnts[cntindex] + 1
+    y <- c(y, cnts[cntindex])
+}
+plot(warriors$Ht_inches, y,
+     xlim = c(69, 87),
+     ylim = c(0, 10),
+     type="n", 
+     main="Player heights", xlab="inches", ylab="count")
+points(warriors$Ht_inches, y, pch=21, col=NA, bg="#999999")
+
+
+# Dots for all players
+htrange <- range(players$Ht_inches)
+cnts <- rep(0, htrange[2]-htrange[1]+1)
+y <- c()
+for (i in 1:dim(players)[1]) {
+    cntindex <- players$Ht_inches[i] - htrange[1] + 1
+    cnts[cntindex] <- cnts[cntindex] + 1
+    y <- c(y, cnts[cntindex])
+}
+
+par(las=1)
+plot(players$Ht_inches, y,
+     xlim = htrange,
+     ylim = c(0, max(cnts)),
+     type="n", 
+     main="Player heights", xlab="inches", ylab="count")
+points(players$Ht_inches, y, pch=21, col=NA, bg="#999999")
+
+# Bar chart version of counts
+barplot(cnts, names.arg=69:87, 
+        main="Player heights", 
+        xlab="inches", ylab="count", border=NA, las=1)
+
+
+# Histogram
+hist(players$Ht_inches, 
+     breaks=seq(65, 90, 1),
+     xlab="inches",
+     main="NBA Player Heights")
+
+# Vary bin sizes
+par(mfrow=c(1,3), mar=c(3,3,3,3))
+hist(players$Ht_inches, main="NBA Player Heights", xlab="inches", breaks=seq(65, 90, 1))
+hist(players$Ht_inches, main="NBA Player Heights", xlab="inches", breaks=seq(65, 90, 2))
+hist(players$Ht_inches, main="NBA Player Heights", xlab="inches", breaks=seq(65, 90, 5))
+
+
+