Load data
#df$labels <- c(1,1,1,1,1,1,2,2,2,2,2,2,3,3,3,3,3,3,4,4,4,4,4,4,5)
df$labels <- c(rep("1", 4), rep("2", 4), rep("3", 4), rep("4", 4), "5")
Error in `$<-.data.frame`(`*tmp*`, labels, value = c("1", "1", "1", "1", :
replacement has 17 rows, data has 25
Summary stats, visualization
ggboxplot(df, x="labels", y="cola")

# Pairwise comparisons
tukey_hsd(df, rte ~ as.factor(labels))
tukey_hsd(df, cola ~ as.factor(labels))
tukey_hsd(df, mrpc ~ as.factor(labels))
tukey_hsd(df, sst2 ~ as.factor(labels))
tukey_hsd(df, qnli ~ as.factor(labels))
tukey_hsd(df, qqp ~ as.factor(labels))
Still use regression vs. ctrl setting
model <- glm(formula, data=df[x_y_features])
Error in y - mu : non-numeric argument to binary operator
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