Answers for "anova() function in r interpretation"

anova() function in r interpretation

raw_pvalue <- numeric(length = length(1:4))
for (i in (1:4)) {
  raw_pvalue[i] <- t.test(dat[, i] ~ dat$Species,
    paired = FALSE,
    alternative = "two.sided"
  )$p.value
}

df <- data.frame(
  Variable = names(dat[, 1:4]),
  raw_pvalue = round(raw_pvalue, 3)
)

df$Bonferroni <-
  p.adjust(df$raw_pvalue,
    method = "bonferroni"
  )
df$BH <-
  p.adjust(df$raw_pvalue,
    method = "BH"
  )
df$Holm <-
  p.adjust(df$raw_pvalue,
    method = "holm"
  )
df$Hochberg <-
  p.adjust(df$raw_pvalue,
    method = "hochberg"
  )
df$Hommel <-
  p.adjust(df$raw_pvalue,
    method = "hommel"
  )
df$BY <-
  round(p.adjust(df$raw_pvalue,
    method = "BY"
  ), 3)
df

anova() function in r interpretation

dat <- iris

# Edit from here
x <- which(names(dat) == "Species") # name of grouping variable
y <- which(names(dat) == "Sepal.Length" # names of variables to test
| names(dat) == "Sepal.Width" |
  names(dat) == "Petal.Length" |
  names(dat) == "Petal.Width")
method1 <- "anova" # one of "anova" or "kruskal.test"
method2 <- "t.test" # one of "wilcox.test" or "t.test"
my_comparisons <- list(c("setosa", "versicolor"), c("setosa", "virginica"), c("versicolor", "virginica")) # comparisons for post-hoc tests
# Edit until here


# Edit at your own risk
for (i in y) {
  for (j in x) {
    p <- ggboxplot(dat,
      x = colnames(dat[j]), y = colnames(dat[i]),
      color = colnames(dat[j]),
      legend = "none",
      palette = "npg",
      add = "jitter"
    )
    print(
      p + stat_compare_means(aes(label = paste0(..method.., ", p-value = ", ..p.format..)),
        method = method1, label.y = max(dat[, i], na.rm = TRUE)
      )
      + stat_compare_means(comparisons = my_comparisons, method = method2, label = "p.format") # remove if p-value of ANOVA or Kruskal-Wallis test >= alpha
    )
  }
}

anova() function in r interpretation

library(ggpubr)

# Edit from here #
x <- which(names(dat) == "Species") # name of grouping variable
y <- which(names(dat) == "Sepal.Length" # names of variables to test
| names(dat) == "Sepal.Width" |
  names(dat) == "Petal.Length" |
  names(dat) == "Petal.Width")
method <- "t.test" # one of "wilcox.test" or "t.test"
paired <- FALSE # if paired make sure that in the dataframe you have first all individuals at T1, then all individuals again at T2
# Edit until here


# Edit at your own risk
for (i in y) {
  for (j in x) {
    ifelse(paired == TRUE,
      p <- ggpaired(dat,
        x = colnames(dat[j]), y = colnames(dat[i]),
        color = colnames(dat[j]), line.color = "gray", line.size = 0.4,
        palette = "npg",
        legend = "none",
        xlab = colnames(dat[j]),
        ylab = colnames(dat[i]),
        add = "jitter"
      ),
      p <- ggboxplot(dat,
        x = colnames(dat[j]), y = colnames(dat[i]),
        color = colnames(dat[j]),
        palette = "npg",
        legend = "none",
        add = "jitter"
      )
    )
    #  Add p-value
    print(p + stat_compare_means(aes(label = paste0(..method.., ", p-value = ", ..p.format..)),
      method = method,
      paired = paired,
      # group.by = NULL,
      ref.group = NULL
    ))
  }
}

anova() function in r interpretation

t_table <- function(data, dvs, iv,
                    var_equal = TRUE,
                    p_adj = "none",
                    alpha = 0.05,
                    paired = FALSE,
                    wilcoxon = FALSE) {
  if (!inherits(data, "data.frame")) {
    stop("data must be a data.frame")
  }

  if (!all(c(dvs, iv) %in% names(data))) {
    stop("at least one column given in dvs and iv are not in the data")
  }

  if (!all(sapply(data[, dvs], is.numeric))) {
    stop("all dvs must be numeric")
  }

  if (length(unique(na.omit(data[[iv]]))) != 2) {
    stop("independent variable must only have two unique values")
  }

  out <- lapply(dvs, function(x) {
    if (paired == FALSE & wilcoxon == FALSE) {
      tres <- t.test(data[[x]] ~ data[[iv]], var.equal = var_equal)
    } else if (paired == FALSE & wilcoxon == TRUE) {
      tres <- wilcox.test(data[[x]] ~ data[[iv]])
    } else if (paired == TRUE & wilcoxon == FALSE) {
      tres <- t.test(data[[x]] ~ data[[iv]],
        var.equal = var_equal,
        paired = TRUE
      )
    } else {
      tres <- wilcox.test(data[[x]] ~ data[[iv]],
        paired = TRUE
      )
    }

    c(
      p_value = tres$p.value
    )
  })

  out <- as.data.frame(do.call(rbind, out))
  out <- cbind(variable = dvs, out)
  names(out) <- gsub("[^0-9A-Za-z_]", "", names(out))

  out$p_value <- ifelse(out$p_value < 0.001,
    "<0.001",
    round(p.adjust(out$p_value, p_adj), 3)
  )
  out$conclusion <- ifelse(out$p_value < alpha,
    paste0("Reject H0 at ", alpha * 100, "%"),
    paste0("Do not reject H0 at ", alpha * 100, "%")
  )

  return(out)
}

anova() function in r interpretation

library(ggstatsplot)

# Comparison between sexes

# edit from here
x <- "sex"
cols <- 3:6 # the 4 continuous dependent variables
type <- "parametric" # given the large number of observations, we use the parametric version
paired <- FALSE # FALSE for independent samples, TRUE for paired samples
# edit until here

# edit at your own risk
plotlist <-
  purrr::pmap(
    .l = list(
      data = list(as_tibble(dat)),
      x = x,
      y = as.list(colnames(dat)[cols]),
      plot.type = "box", # for boxplot
      type = type, # parametric or nonparametric
      pairwise.comparisons = TRUE, # to run post-hoc tests if more than 2 groups
      pairwise.display = "significant", # show only significant differences
      bf.message = FALSE, # remove message about Bayes Factor
      centrality.plotting = FALSE # remove central measure
    ),
    .f = ifelse(paired, # automatically use ggwithinstats if paired samples, ggbetweenstats otherwise
      ggstatsplot::ggwithinstats,
      ggstatsplot::ggbetweenstats
    )
  )

# print all plots together with statistical results
for (i in 1:length(plotlist)) {
  print(plotlist[[i]] +
    labs(caption = NULL)) # remove caption
}

anova() function in r interpretation

dat <- iris

# remove one level to have only two groups
dat <- subset(dat, Species != "setosa")
dat$Species <- factor(dat$Species)

# boxplots and t-tests for the 4 variables at once
for (i in 1:4) { # variables to compare are variables 1 to 4
  boxplot(dat[, i] ~ dat$Species, # draw boxplots by group
    ylab = names(dat[i]), # rename y-axis with variable's name
    xlab = "Species"
  )
  print(t.test(dat[, i] ~ dat$Species)) # print results of t-test
}