# FC721 Introduction to R Programming
# Part2: Working with Data in R
#-----------------------------------

library(tidyverse)

# Load the data
# There is a "regular" R function to read csv files: read.csv()
nhanes_r <- read.csv("nhanes_2018_df.csv")

nhanes_tidy <- read_csv("nhanes_2018_df.csv")

# Check the structure of the data
str(nhanes_r)
str(nhanes_tidy)
# nhanes_tidy is a tibble, which is a modern version of a data frame
# tibbles are part of the tidyverse, and are more user-friendly than data frames
# when you work with tibbles, you can use glimpse() instead of str()
glimpse(nhanes_tidy)

# When you use str() and glimpse() functions, you can see the data type of each column
# Make sure that the data type of each column is correct

# <dbl> (num) stands for double, which is a numeric data type
# <int> (int) stands for integer
# <chr> (chr) stands for character, which is a text data type
# <fct> (factor) stands for factor, which is a categorical data type
# <lgl> (logi) stands for logical, which is a boolean data type
# and some others

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

# Next we need to explore our dataset to understand its structure
# We can use the following functions to explore the data:
# head(), tail(), summary(), colnames(), unique(), table()

# head() function shows the first 6 rows of the data
head(nhanes_tidy)
# To see more or less rows, you can specify the number of rows you want to see:
head(nhanes_tidy, 10)

# To see the last 6 rows of the data, you can use tail() function
tail(nhanes_tidy)

# summary() function provides a quick summary of the data
summary(nhanes_tidy)

# Make sure to carefully examine the output of the summary() function
# It provides information about the minimum, 1st quartile, median, mean, 3rd quartile, and maximum values for each column
# It also provides the number of missing values in each column
# Check if the minimum and maximum values make sense for each column
# Check if the number of missing values is acceptable
# If there are any issues, you may need to clean the data

# For the columns that are character, use unique() function to see unique values:
unique(nhanes_tidy$race)

# For the columns that contain categorical values, use table() function to see the frequency of each category:
table(nhanes_tidy$race)

# Note: categorical variable may be stored as a character or numeric data type
# You may need to convert it to a factor data type using as.factor() function:
nhanes_tidy$race <- as.factor(nhanes_tidy$race)

# Let's check the structure of the data again
glimpse(nhanes_tidy)
# Notice the difference in the output of the summary() function for the race column:
# It now shows the frequency of each category:
summary(nhanes_tidy)


#-----------------------------------
#   |> and  %>% "pipes"
#-----------------------------------

# Native R's pipe symbol |> was introduced in 4.1.0
# It allows to send the output of one function as an input object to the next function:
nhanes_tidy |> head()

# You may see another "pipe" symbol used very frequently.
# %>% "pipe" symbol was introduced in magrittr package long time ago and 
# then it was adopted by other packages in tidyverse library and some other R packages
nhanes_tidy %>% head()


#-----------------------------------
#  Data Manipulation
#-----------------------------------

# 1. Select columns you want to keep
# 2. Filter rows you want to keep
# 3. Mutate columns to create new columns or modify existing ones
# 4. Group data (if needed)
# 5. Summarize data 


# 1. Select columns you want to keep
# Use select() function.
# You can specify the columns you want to keep by their names and use
# comma to select specific columns or colon to select a range of columns:
df1 <- nhanes_tidy |> select(id, gender, age, bp_systolic: weight)
# This is equivalent to:
df1 <- select(nhanes_tidy, id, gender, age, bp_systolic: weight)

# 2. Filter rows you want to keep
df2 <- df1 |> filter ( ! is.na(bp_systolic) & ! is.na(bp_diastolic) )

# Useful functions for filtering:
# is.na() - checks if a value is missing
# %in% - checks if a value is in a list of values
# between() - checks if a value is between two values
# == - checks if a value is equal to another value
# != - checks if a value is not equal to another value
# >, <, >=, <= - checks if a value is greater than, less than, greater than or equal to, less than or equal to another value
# & - logical AND
# | - logical OR

nhanes_tidy |> filter ( between(age, 18, 65) ) |> head()
nhanes_tidy |> filter ( race %in% c("Non-Hispanic Black", "Mexican American") ) |> head()

# 3. Mutate columns to create new columns or modify existing ones
df3 <- df2 |> mutate( height_ft = height / 30.48)
head(df3)
summary(df3$height_ft)

# Useful functions for mutating:
# if_else() - conditional statement
# case_when() - multiple conditional statements
df2 |> mutate(age_group = if_else(age < 18, "Child", "Adult")) |> head()


# Instead of doing this in 3 separate steps, you can chain them together using pipe symbols:
df <- nhanes_tidy |> 
  select(id,gender, age, bp_systolic: weight) |>
  filter( ! is.na(bp_systolic) & ! is.na(bp_diastolic) ) |>
  mutate( height_ft = height / 30.48 )


# 4. - 5. Group (if needed) and summarize data:
nhanes_tidy |> 
  group_by(gender) |>
  summarize(mean_age = mean(age), 
            mean_weight = mean(weight, na.rm = TRUE),
            mean_height = mean(height, na.rm = TRUE))

# You can also use a new ".by" argument in summarize() function:
nhanes_tidy |> 
  summarize(mean_age = mean(age), 
            mean_weight = mean(weight, na.rm = TRUE),
            mean_height = mean(height, na.rm = TRUE),
            .by = c(gender, race))


# use n() function to count the number of rows in each group:

nhanes_tidy |> 
  group_by(race) |>
  summarize(N= n(),
            mean_age = mean(age))


#-----------------------------------
# Exercise
#-----------------------------------

# 1. Using nhanes_tidy dataset select the columns id, gender and age
# 2. Filter the rows where age is greater than 18
# 3. Mutate the dataset to create a new column bmi: bmi = weight / (height / 100)^2
# 4. Save the result to a new dataset called nhanes_adults
# 5. Using write_csv() function save the resulting dataset to a new csv file called nhanes_adults.csv

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



#-----------------------------------
# A few more useful functions
#-----------------------------------

# 1. arrange() function is used to sort the data by one or more columns
nhanes_tidy |> arrange(age) |> head()

# 2. distinct() function is used to remove duplicate rows
nhanes_tidy |> distinct() |> head()

# 3. rename() function is used to rename columns
nhanes_tidy |> rename(Height = height) |> head()

# 4. count() function is used to count the number of rows for each group
nhanes_tidy |> count(race)


#-----------------------------------
# You may find table1 package useful for creating table1 summaries
#-----------------------------------
install.packages("table1")
library(table1)

table1( ~ age + bp_systolic + bp_diastolic + height + weight | race, data = nhanes_tidy)