Visualize demographics for scans

Signed-off-by: Zarrar Shehzad <czarrar@gmail.com>

nki/02_qc/20_choose_scans.R

@@ -0,0 +1,67 @@
+#' This script does some quick comparisons of the scans following QC
+#' 
+
+#+ setup, include=FALSE
+library(ggplot2)
+df <- read.csv("../subinfo/qc-2_usable_scans.csv")
+df <- df[,colnames(df)!="subdir"] # remove the path column
+
+
+###
+# Examine 2 of the scans
+###
+
+#' # Two Scans
+
+#' Here we examine subjects who have 645 and 1400ms TR scans that pass QC.
+#' We look at the distribution of ages and the breakdown of sex and handedness.
+#+ two-scans-subset
+sdf <- subset(df, short==1 & medium==1)
+
+#' Age appears to be bimodally distributed, is this an issue?
+#+ two-scans-age
+ggplot(sdf, aes(Age)) + geom_histogram(binwidth=5) + ggtitle("Age for Participants with 2 Scans")
+
+#' Note there is one participant who doesn't have gender recorded.
+#+ two-scans-sex
+table(sdf$Sex)
+ggplot(sdf, aes(Sex)) + geom_bar() + ggtitle("Sex for Participants with 2 Scans")
+
+#' Woot, majority of people are right-handed.
+#+ two-scans-handedness
+table(sdf$Handedness)
+ggplot(sdf, aes(Handedness)) + geom_bar() + ggtitle("Handedness for Participants with 2 Scans")
+
+
+###
+# Examine all three scans
+###
+
+#' # Three Scans
+
+#' Here I will be seeing how the distribution of the subjects from above
+#' will differ with the removal of subjects who don't have a 3rd scan.
+#' Where color is shown, red means subjects with only two scans, while
+#' blue means subjects with all three scans usable.
+
+#' Subjects that would be lost if looking at three scans are in red (shown first)
+#' Only subjects with three usable scans are shown second.
+#+ three-scans-age
+ggplot(sdf, aes(Age, fill=factor(all))) + geom_histogram(binwidth=5) + ggtitle("Age for Participants with 2/3 Scans")
+ggplot(sdf[sdf$all==3,], aes(Age)) + geom_histogram(binwidth=5) + ggtitle("Age for Participants with 3 Scans")
+
+#' The first table output is for 2 scans and the second is for 3 scans.
+#+ three-scans-sex
+table(sdf$Sex)
+table(sdf$Sex[sdf$all==3])
+ggplot(sdf, aes(Sex, fill=factor(all))) + geom_bar() + ggtitle("Sex for Participants with 2/3 Scans")
+ggplot(sdf[sdf$all==3,], aes(Sex)) + geom_bar() + ggtitle("Sex for Participants with 3 Scans")
+
+#' The first table output is for 2 scans and the second is for 3 scans.
+#' Woot, majority of people are still right-handed.
+#+ three-scans-handedness
+table(sdf$Handedness)
+table(sdf$Handedness[sdf$all==3])
+ggplot(sdf, aes(Handedness, fill=factor(all))) + geom_bar() + ggtitle("Handedness for Participants with 2/3 Scans")
+ggplot(sdf[sdf$all==3,], aes(Handedness)) + geom_bar() + ggtitle("Handedness for Participants with 3 Scans")
+