Question

I’ve come up against a wall, and believe I am missing something quite simple.

I have a list that contains results from multiple wilcoxon tests (with Bonferroni corrections) and would like to produce a similar-looking table to the output of the Dunnett’s test using multcomp (i.e., with no row numbers showing up and nicely spaced).

When I print it as a dataframe, the row numbers show up and the text values in the columns are right-justified.

The list is created by:

for (i in 2:length(split.set)) {
  wrs.mod <- suppressWarnings(wilcox.test(split.set[[1]]$VALUE, split.set[[i]]$VALUE))
  stn.results[i - 1] <- as.character(unique(split.set[[i]]$TREATMENT))
  stat.results[i - 1] <- as.numeric(wrs.mod$statistic)
  p.results[i - 1] <- signif(wrs.mod$p.value, 3)
  if (wrs.mod$p.value < 0.05/(length(split.set) - 1)) sig.results[i - 1] <- "*" else sig.results[i - 1] <- NA
}
wrs.results <- list(Treatment = stn.results, Statistic = stat.results, p = p.results, Significant = sig.results)

How do I format this into a decent looking table for printing?

Answer 1

Instead of formatting it as a list, format it as a data.frame and print that:

wrs.results <- data.frame(Treatment = stn.results, Statistic = stat.results, Pvalue = p.results, Significant = sig.results)
print(wrs.results)

You can customize the structure (and therefore the look) of the data frame in a couple of ways. You might prefer treatment to be row names:

wrs.results <- data.frame(row.names = stn.results, Statistic = stat.results, Pvalue = p.results, Significant = sig.results)

or customize the column names:

colnames(wrs.results) = c("Treatment", "Statistic", "P(>|t|)", "Significance")