#############
R-code notes compilation for  Week 5 February 15th
Compiled by: Deepta Basak and Bikram Halder B.Math(hons.) 1st year
################


# R categorical data
# Factors and Levels
# - used to plot categorical data in R

# See the following packages in the book - R for Data Science
# - dplyr (https://r4ds.had.co.nz/transform.html#dplyr-basics)
# - forcats (https://r4ds.had.co.nz/factors.html#prerequisites-9)


# Keys: How base-R handles categorical data or variables?


# Categorical Variables - Qualitative variables
# ,i.e., those which cannot meaningfully be expressed in numbers.
# Eg.: Clothes, Colour.
# -> Base-R deals with them through the use of factors.
# -> factors are useful in Statistical Modeling and Plotting data.

# -> Packages - tidyverse- dplyr, tidyr, forcats, readr
# help in dealing with factors.

# factor(x)
# - Store as vector  of integers
# - Displayed as characters

####
xc = c("June", "July", "August", "September",
       "August", "July","July","August" )
#create a vector of data
factorxc  = factor(xc)
#store the vector as a factor using factor()
factorxc
#factors in R are stored as a vector of integers but they correspond to a character string for display
#Levels are the unique set of values taken by as.character() [this is the default assignment] 
# factor levels are always characters
###



# xn - numeric data as factors
xn <- c(1, 2, 2, 3, 1, 2, 3, 3, 1, 2, 3, 3, 1)
# Create a vector of data
factorxn <- factor(xn)
# Store it as a factor using factor()
factorxn
# Note factor levels are always characters



# Both numeric and character data can be stored as factors
mean(xn)
# But, factor levels are always characters
mean(factorxn)
# Since, argument is not numeric, so will return NA

# use levels() function to convert them to original numeric value
mean(as.numeric(levels(factorxn)[factorxn]))


# xc - character data as factors
xc <- c(
  "June", "July", "August", "September",
  "August", "July", "July", "August"
)
factorxc <- factor(xc)
factorxc




#####
table(factorxc)
# table() gives frequency of each level

# ordering is w.r.t. to as.character(.)  and not related to the order in the vector.
months <- factor(xc,
                 levels = c(
                   "Garbage", "January", "February",
                   "March", "April", "May", "June", "July", "August",
                   "September", "October", "November", "December"
                 ),
                 ordered = TRUE
)
#we can specify the levels and the order that we want
months
# Here, in our specification, levels contain elements not present in months
months[3] < months[4]

# When factor is created, all it's levels are stored with the factor
table(months)
# - table() will display  ALL the levels specified and respective counts
# - occur when subsetting a factor

factor(months)
# retains only the levels present without affecting the ordering (i.e and maintains their order.)
table(factor(months))
# Gives the count of ONLY the levels present in the factor



#####
x <- round(1000 * runif(10))
x

# cut() function is used to convert a numeric variable into a factor
# Syntax: cut(numeric data, breaks)
xfactor <- cut(x, 4)
# breaks
# - arbitary by default
# - specifies how range of numbers will be converted to factor values
xfactor
table(xfactor)
#displays the count of each level that was created using cut()



cut(x, 3, labels = c("L", "M", "H"))
# label() - used to specify the levels of the factors

# Nicer set of labels with pretty()
xpfactor <- cut(x, pretty(x, 4))
xpfactor
# but may or may not provide more levels than specified
table(xpfactor)
#displays the count of each level created with pretty()


xqfactor <- cut(x, quantile(x, probs = seq(0, 1, 0.25)))
# Produce factors based on percentiles of your data

table(xqfactor)
# observations are distributed "equally" in each level





everyday <- seq(
  from = as.Date("2021-1-1"),
  to = as.Date("2021-12-31"),
  by = "day"
)
##### Create factors from dates/times
# strptime(), strftime() with factors and levels would be very useful for extracting information with ordering from a scraped dataset.

cmonth <- format(everyday, "%b")
#format() can be used to extract the month from each date using '%b'
# Extract month from each day
head(cmonth, 3)

df <- as.data.frame(table(cmonth))
# table()
# - tabulates values in each month
# - ordering in alphabetcal
# as.data.frame() stores the output as dataframe
names(df) <- c("Month", "Freq")
#changing the names of the variables of the dataframe d
df


months <- factor(cmonth,
                 levels = unique(cmonth), ordered = TRUE
)
# unique(): returns the unique values in the order that are encountered
# stored months as factors
df2 <- as.data.frame(table(months))
names(df2) <- c("Month", "Freq")
# specified levels & ordering using the unique()
df2



everyday <- seq(
  from = as.Date("2021-1-1"), to = as.Date("2021-12-31"),
  by = "day"
)
## vector of all Dates of the year 2021

wks <- cut(everyday, breaks = "week")
# levels - 53 dates of each week of the year 2021
wks

qtrs <- cut(everyday, "3 months",
            labels = paste("Q", 1:4, sep = "")
)
# levels - 4 quarters of the year 2021
qtrs






###
decdf <- read.csv(
  file = "https://www.isibang.ac.in/~athreya/Teaching/ISCD/Master.csv",
  header = TRUE
)
# This is the deceased data from the Government of Karnataka COVID-19 Bulletin.
# The https://www.isibang.ac.in/~athreya/Teaching/ISCD/Master.csv file contains this collation.

decdf$Month <- months(as.Date(decdf$MB.Date))
# Use months() to extract months of reporting date
data <- as.data.frame(table(decdf$Month))
names(data) <- c("name", "val")
# Use table() to compute reported cases across months
# change names of dataframe
data


library(tidyverse)
# Loading tidyverse package
ggplot(data = data, aes(x = name, y = val, fill = name)) +
  geom_bar(stat = "identity", alpha = .6, width = .4) +
  coord_flip() +
  scale_fill_viridis_d() +
  xlab("") +
  theme_bw()
# Note:
# - ggplot uses ordering alphabetically
# - treats Month as factor and default level ordering



####
## try to plot below in order of frequency values.
decdf <- read.csv(
  file = "https://www.isibang.ac.in/~athreya/Teaching/ISCD/Master.csv",
  header = TRUE
)
decdf$Month <- months(as.Date(decdf$MB.Date))
data <- as.data.frame(table(decdf$Month))
names(data) <- c("name", "val")
#create data frame
data <- arrange(data, val)
# Note: We arranged our dataframe to preferred ordering.

library(tidyverse)
ggplot(data = data, aes(x = name, y = val, fill = name)) +
  geom_bar(stat = "identity", alpha = .6, width = .4) +
  scale_fill_viridis_d() +
  coord_flip() +
  xlab("") +
  theme_bw()
# ggplot() takes into account ordering from the factor given by its levels
# & NOT as we see in the data-frame
# Inspite of arranging our dataframe, the plot is still the same as before!


######
# Same code as before - some correction to get required result
## try to plot below in order of frequency values.

library(tidyverse)
decdf <- read.csv(file = "../Master.csv", header = TRUE)
decdf$Month <- months(as.Date(decdf$MB.Date))
data <- as.data.frame(table(decdf$Month))
names(data) <- c("name", "val")
data <- arrange(data, val)

data$name <- factor(data$name, levels = data$name)
# Reorder the levels according to the values

ggplot(data = data, aes(x = name, y = val, fill = name)) +
  geom_bar(stat = "identity", alpha = .6, width = .4) +
  coord_flip() +
  scale_fill_viridis_d() +
  xlab("") +
  theme_bw()
# ggplot() obliges the order now


######
#using dplyr and forcats
decdf <- read.csv(
  file = "https://www.isibang.ac.in/~athreya/Teaching/ISCD/Master.csv",
  header = TRUE
)
decdf$Month <- months(as.Date(decdf$MB.Date))
#create data frame

library(tidyverse)
ggplot(
  data = decdf,
  aes(
    x = fct_infreq(Month),
    fill = cut(Age.In.Years, pretty(Age.In.Years, 4))
  )
) +
  geom_bar(stat = "count", alpha = .6, width = .4) +
  scale_fill_viridis_d() +
  xlab("") +
  coord_flip() +
  theme_bw()

# With dplyr(R package): Order months according to reported deaths
# With forcats(R package): fct_infreq

###
library(dplyr)
library(forcats)
ggplot(
  decdf,
  aes(
    x = fct_infreq(Month),
    fill = cut(Age.In.Years, pretty(Age.In.Years, 4)) # cut() is used to add Age.In.Years bin layer
  )
) +
  labs(fill = "Age") +
  geom_bar(stat = "count", width = .4) +
  scale_fill_viridis_d() +
  coord_flip() +
  theme_bw() +
  ggtitle("Deceased Month Count- Age Stacked") +
  ylab("Count of Deceased") +
  xlab("Month") +
  theme_bw() +
  theme(
    plot.title = element_text(
      color = "#e95462",
      size = 14,
      face = "bold",
      hjust = 0.5
    ),
    axis.title.x = element_text(
      color = "#e95462",
      size = 14, vjust = 0.5, face = "bold"
    ),
    axis.title.y = element_text(
      color = "#e95462",
      size = 14, face = "bold"
    )
  )
# forcats used to create ordered factors.
#fill command used to add age as layer.
#labeling of axis, title and ticks discussed above.
# check other themes in ggplot
#cut() from base R was used to add Age bin layer
## Viridis hexcodes from https://waldyrious.net/viridis-palette-generator/