Checking R code and this debug error: unexpected '}' in "}"

Question

I just need some feedback on a very simple code I made, it's also saying Error: unexpected '}' in "}", and I wonder if this is a unicode issue? Any feedback would be appreciated!!

My data is here: joint.sexdiff.csv

I'm a really really new coder, and R is the one I want to use for future stats, so I'm constantly learning and appreciate all the feedback!

# Statistical test pipeline when trying to compare two groups ----
# Description: When trying to compare two groups, t-test is often used. Unpaired two-samples t-test can be used only under certain conditions:
# (1) when the two groups of samples (A and B), being compared, are normally distributed
# This can be checked using Shapiro-Wilk test
# (2) When the variances of the two groups are equal
# This can be checked using F-test

# Given a sample dataframe, a pipeline will be used to automatically determine whether a parametric or non-parametric test should be used, and get the p-value

# Packages to install and call
install.packages("dplyr")
install.packages("ggplot2")
library(dplyr)
library(ggplot2)

# Sample data named "joint.sexdiff.csv"
data1<-read.csv("*path*")


# Some descriptive stats
group_by(data1, Sex) %>%
  summarise(
    count=n(),
    mean=mean(JS.W, na.rm=TRUE)
  )

# Data visualization using ggplot2
bplot<-ggplot(
  data1, aes(x=Sex, y=JS.W, colour=Sex))+
    geom_boxplot(outlier.shape = NA)+
    geom_jitter()+
    ylab("Joint space width (mm)")+
    xlab("")+
  ggtitle("Male vs Female joint space width (JSW)")
bplot
  
# Now we have the data we need, we can run a two group comparison
# Use Shapiro-Wilk test to test if the joint space widths fall within a normal distribution, if p>0.05, we can assume normality
# If the distribution is not normal, we need to use a non-parametric test

# Non-parametric unpaired two-samples Wilcoxon test is used to compare two independent groups of samples that are not normally distributed
# http://www.sthda.com/english/wiki/unpaired-two-samples-wilcoxon-test-in-r

# The process is as follows:
# Shapiro test -> if normal distribution or p>0.05, do F-test -> if variances are equal or p>0.05, var.equal=TRUE will be assumed on the t-test
# -> if variances are unequal or p<0.05, var.equal=FALSE will be assumed
# IMPORTANT -- your t-test p-values will change depending on whether you assume the variances to be equal or not
# Shapiro -> if distribution is not normally distributed or p<0.05, use Wilcoxon test

# The test blocks of code will check your data and run an independent samples t-test or Wilcoxon test depending on whether sample is normally distributed and variances are normal

# Output = p.value

shapJSW<-shapiro.test(data1$JS.W)
if (shapJSW$p.value>0.05){
  ftestJSW<-var.test(JS.W ~ Sex, data = data1)
  if (ftestJSW$p.value>0.05){
    ttestJSW<-t.test(JS.W ~ Sex, data = data1, var.equal = TRUE)
    print(ttestJSW$p.value)
  } else {
    ttestJSW<-t.test(JS.W ~ Sex, data = data1, var.equal = FALSE)
    print(ttestJSW$p.value)
} else {
  wtestJSW<-wilcox.test(JS.W ~ Sex, data = data1)
  print(wtestJSW)
  wtestJSW<-wilcox.test(JS.W ~ Sex, data = data1)
  print(wtestJSW)
}
}

Sprinkle some of your wisdom to me! :) <3

Answer 1

The error is a missing } after printing the p-value of the t-test with unequal variances. The testing part can be re-written like this:

test_sex <- function(d) {
  if(shapiro.test(d$JS.W)$p.value>0.05){
    equal_var = var.test(JS.W~Sex,data=d)$p.value>0.05
    ttestJSW = t.test(JS.W ~ Sex, data = d, var.equal = equal_var )
    return(list(dist = "normal", test="t-test", equal_var=equal_var, test_result = ttestJSW))
  }
  list(dist = "non-normal", test="wilcox", test_result = wilcox.test(JS.W ~ Sex, data = d))
}

Usage:

test_sex(data1)

Output:

$dist
[1] "non-normal"

$test
[1] "wilcox"

$test_result

    Wilcoxon rank sum exact test

data:  JS.W by Sex
W = 100, p-value = 0.01075
alternative hypothesis: true location shift is not equal to 0