# let's create vector with mean 3 and standard deviation 1
# normally distributed in interval 0-5
# 100 observations

set.seed(123)
x <- rnorm(100, mean = 3, sd = 1)
hist(x, breaks = 20, col = "lightblue", main = "Histogram of x", xlab = "x")

# now let's make for loop to create 72*100 observations
# and store them in 24 vectors a - x
# each vector will have different mean, starting from 0.5 and going up by 0.1

for (i in 1:24) {
  assign(paste0("a", letters[i]), rnorm(100, mean = 0.5 + 0.1*i, sd = 1))
}

# next let's create 24 vectors A - X

for (i in 1:24) {
  assign(paste0("A", letters[i]), rnorm(100, mean = 3 + 0.1*i, sd = 1))
}

# now let's put all the vectors into one data frame

df <- data.frame()

for (i in 1:24) {
  df <- rbind(df, data.frame(condition = paste0("a", letters[i]), value = get(paste0("a", letters[i]))))
}

for (i in 1:24) {
  df <- rbind(df, data.frame(condition = paste0("A", letters[i]), value = get(paste0("A", letters[i]))))
}

# let's add x as a condition to the data frame

df <- rbind(df, data.frame(condition = "x", value = x))

# now let's make a boxplot

boxplot(value ~ condition, data = df, col = "lightblue", main = "Boxplot of x and aw", xlab = "condition", ylab = "value")

# and load standard lme4 package
# plus lmerTest for p-values

# library(lme4)
library(lmerTest)

# subset x as baseline and aa, Aa as conditions
# via tidyverse

library(tidyverse)

df2 <- df %>%
  filter(condition %in% c("x", "aa", "Aa"))

# ggplot histogram for df2

library(ggplot2)

ggplot(df2, aes(x = value, fill = condition)) +
  geom_histogram(bins = 20, alpha = 0.5, position = "identity") +
  labs(title = "Histogram of x, aa and Aa", x = "value", y = "count") +
  theme_minimal()

# let's compute linear model with x as baseline

as.factor(df2$condition) -> df2$condition

df2$condition <- relevel(df2$condition, ref = "x")

model <- lmer(value ~ condition + (1|condition), data = df2, REML = FALSE)

summary(model)