functions-homework.R

#PSYC 259 Homework 4 - Writing functions
#For full credit, answer at least 6/8 questions

#List names of students collaborating with: 

### SETUP: RUN THIS BEFORE STARTING ----------
library(tidyverse)
set.seed(1)
id <- rep("name", 30)
x <- runif(30, 0, 10)
y <- runif(30, 0, 10)
z <- runif(30, 0, 10)
ds <- tibble(id, x, y, z)

### Question 1 ---------- 

#Vectors x, y, and z contain random numbers between 1 and 10. 
#Write a function called "limit_replace" that will replace values less than 2 or greater than 8 with NA
#Then, run the function on x and save the results to a new vector "x_replace" to show it worked


### Question 2 ---------- 

#Make a new version of limit_replace that asks for arguments for a lower bounary and an upper boundary
  #so that they can be customized (instead of being hard coded as 2 and 8)
#Run the function on vector y with boundaries 4 and 6, saving the results to a new vector "y_replace"


### Question 3 ----------

#Write a function called "plus_minus_SD" that can take one of the vectors (x/y/z) as input
  #and "num_of_SDs" as an input and returns the boundaries +/- num_of_SDs around the mean. 
#plus_minus_SD(x, 1) would give +/- 1SD around the mean of x, plus_minus_SD(y, 2) would give +/- 2SDs around the mean 
#Make num_of_SDs default to 1
#run the new function on x, y, and z with 1 SD


### Question 4 ----------

#Write an another new version of limit_replace
#This time, make the upper and lower boundaries optional arguments
#If they are not given, use +/- 1 SD as the boundaries (from your plus_minus_SD function)
#Apply the function to each column in ds, and save the results to a new tibble called "ds_replace"



### Question 5 ----------

#Add a "stopifnot" command to your limit_replace function to make sure it only runs on numeric variables
#Try running it on a non-numeric input (like "id") to make sure it gives you an error



### Question 6 ----------

#What other requirements on the input do you need to make the function work correctly?
#Add another stopifnot to enforce one more requirement



### Question 7 ----------

#Clear out your workspace and load the built-in diamonds dataset by running the lines below
#RUN THIS CODE
rm(list = ls())
library(tidyverse)
ds_diamonds <- diamonds

#Save your two functions to an external file (or files) 
#Then, load your functions from the external files(s)
#Next, run your limit_replace function on all of the numeric columns in the new data set
#and drop any rows with NA, saving it to a new tibble named "ds_trimmed"


### Question 8 ----------

#The code below makes graphs of diamond price grouped by different variables
#Refactor it to make it more efficient using functions and/or iteration
#Don't worry about the order of the plots, just find a more efficient way to make all 6 plots
#Each cut (Premium/Ideal/Good) should have a plot with trimmed and untrimmed data
#The title of each plot should indicate which cut and whether it's all vs. trimmed data

ds_diamonds %>% filter(cut == "Premium") %>% 
  ggplot(aes(x = clarity, y = price)) + 
  geom_boxplot() + 
  ggtitle("Premium, all") + 
  theme_minimal()

ds_diamonds %>% filter(cut == "Ideal") %>% 
  ggplot(aes(x = clarity, y = price)) + 
  geom_boxplot() + 
  ggtitle("Ideal, all") +
  theme_minimal()

ds_diamonds %>% filter(cut == "Good") %>% 
  ggplot(aes(x = clarity, y = price)) + 
  geom_boxplot() + 
  ggtitle("Good, all") +
  theme_minimal()

ds_trimmed %>% filter(cut == "Premium") %>% 
  ggplot(aes(x = clarity, y = price)) + 
  geom_boxplot() + 
  ggtitle("Premium, trimmed") + 
  theme_minimal()

ds_trimmed %>% filter(cut == "Ideal") %>% 
  ggplot(aes(x = clarity, y = price)) + 
  geom_boxplot() + 
  ggtitle("Ideal, trimmed") +
  theme_minimal()

ds_trimmed %>% filter(cut == "Good") %>% 
  ggplot(aes(x = clarity, y = price)) + 
  geom_boxplot() + 
  ggtitle("Good, trimmed") +
  theme_minimal()