data1 <- data.frame(
  "animal" = c(rep("mouse", 10), rep("chipmunk", 10), rep("squirrel", 10)),
  "weight" = c(rnorm(10, mean = 10, sd = 1), rnorm(10, mean = 50, sd = 5),
               rnorm(10, mean = 100, sd = 11))
  )

# --------------------------------------------------------------------
# -------- Tool for Summary of Means Across Categories ---------------
# --------------------------------------------------------------------

# This tool takes a dataset with two variables and calculates the mean
# value of a dependent variable (and 95% CI) for each level of an
# independent variable. It requires three arguments:
# "data" = a dataset containing all the variables for summary)
# "groups" = the name (in quotation marks!) of the column containing the categories
# "values" = the name (in quotation marks!) of the column containing the values for summary
#
# example usage:
#
 data1 <- data.frame(
   "animal" = c(rep("mouse", 10), rep("chipmunk", 10), rep("squirrel", 10)),
   "weight" = c(rnorm(10, mean = 10, sd = 1), rnorm(10, mean = 50, sd = 5),
                rnorm(10, mean = 100, sd = 11))
 )
 sum1 <- summary_table(data = data1, "animal", "weight")

library("dplyr")
library("rlang")
library("lazyeval")

ci <- function(x) {sqrt(sd(x)/length(x)) * 1.96}

summary_table <- function(data, groups, values) {
  data %>% 
    group_by(!!sym(groups)) %>% 
    summarise("Mean" = mean(!!sym(values), na.rm = TRUE), "ConfInt" = ci(!!sym(values)))
}


#################################################
#################################################
#################################################

# --------------------------------------------------------------------

# -----------------    Basic Boxplot    ----------------------------
#
# This is best done with all the components in a single "data" file
# "levels" are the categories for each bar in the barplot (these date are contained within "data")
# "levels" should be subsetted from the main data, e.g., data$levels
# "meanvals" are the actual means (i.e., the bar height) to be plotted
# "meanvals" should be subsetted from the main data, e.g., data$meanvals
# "confints" is the 95% confidence interval for each bar
# "confints" should be subsetted from the main data, e.g., data$meanvals
# "ylabel" should be in QUOTATIONS!
#
# example usage:

data1 <- data.frame("animal" = c("mouse", "chipmunk", "squirrel"), "weight" = c(12, 54, 109), "CI" = c(1.2, 10.1, 24.9))
basic_boxplot(levels = data1$animal, meanvals = data1$weight, confints = data1$CI, ylabel = "weight (g)")

# OR

basic_boxplot(levels = sum1$animal, meanvals = sum1$Mean, confints = sum1$ConfInt, ylabel = "weight (g)")

#####

basic_boxplot <- function(levels, meanvals, confints, ylabel) {
  plotTop <- max(meanvals + confints) * 1.20 # Top of the plot is 20% higher than tallest CI
  barCenters <- barplot(meanvals, names.arg = levels, col="gray", las=1, ylim = c(0, plotTop), ylab = ylabel)
  segments(barCenters, meanvals - confints, barCenters, meanvals + confints, lwd=1) # adding bars
  arrows(barCenters, meanvals - confints, barCenters, meanvals + confints, lwd=1, angle=90, code=3) # adding braces
  axis(side = 1, at = barCenters, labels = FALSE) # adding an x-axis line with tick marks
  abline(h=0) # filling the small gap in the axis
}