• I currently have the flu

• I’m in a lot of pain, and am struggling with things like making food and classwork

• Because of this, I want to know more about the flu

• Even though its morbid, I want to do a project on the Flu, and how many people die a year to it

• I know that’s probably unwise to do while sick with the Flu, but I’m committed

• Data is from the CDC

• Each entry has a year, a state, a death count, and a rate

• Data ranges from 2014 and 2021, with a value from every state every year

• I have a few questions about this data

1. How have flu death totals changed over between 2014 and 2021

2. Make a Choropleth map to show which state has the most average deaths per person

• With both questions asked, let’s get to it

Flu_Data <- read.csv("flu-table.csv")
State_Data <- read.csv("U.S. states population data.csv")

• Now, I need to fix up the data, to make it more usable

Flu_Data$state_name <- state.name[match(Flu_Data$STATE, state.abb)]
Flu_Data$DEATHS <- as.numeric(gsub(",", "", Flu_Data$DEATHS))
Flu_Data$DEATHS <- as.numeric(Flu_Data$DEATHS) 
Flu_Data <- Flu_Data[Flu_Data$YEAR != 2005, ]
State_Data$population <- State_Data$POPULATION..2023.
State_Data <- State_Data[order(State_Data$STATE), ]
State_Data$population <- as.numeric(gsub(",", "", State_Data$population))
State_Data$population <- as.numeric(State_Data$population) 

• With that, let’s start solving some of these problems

1. How have flu death totals changed over between 2014 and 2021

• To start, let’s aggregate this data

• I need the sum of each year in its own data frame

aggregate_data <- aggregate(Flu_Data$DEATHS, list(Flu_Data$YEAR), FUN = sum)
aggregate_data <- aggregate_data[!is.na(aggregate_data$x), ]

• From there, we’re already in a position to create a graph of each year

barplot(aggregate_data$x, 
        names.arg = aggregate_data$Group.1, 
        col = c("lightcoral", "lightsalmon", "lightgoldenrod", "lightgreen",             "lightcyan", "lightblue", "lightpink", "plum2"),
        main = "Bar Plot Example")

• It a bit stagnant, but it’s going down

• We can esepcially see that 2021 is pretty solid

• This gives me hope I wont terribly perish from the flu

1. Make a Choropleth map to show which state has the most average deaths per person

• This is the main reason I’m doing this project

• I’ve wanted to Choropleth map techniques for a while, and this is a perfect exscuse

• Huge thank you to https://r-graphics.org/recipe-miscgraph-choropleth

• To start, lets look get the average data per state

• I’ll also do some formatting to make our lives easier later

aggregate_data <- aggregate(Flu_Data$DEATHS[!is.na(Flu_Data$DEATHS)], list(Flu_Data$state_name[!is.na(Flu_Data$DEATHS)]), FUN = mean)
names(aggregate_data)[1] <- "State"
names(aggregate_data)[2] <- "Cases"
aggregate_data$State <- lapply(aggregate_data$State, tolower)

• Next, we need to take population size into account

• We can use the State_Data frame to get population size, and then devide.

aggregate_data$CasesPerPerson <- aggregate_data$Cases / State_Data$population

• Now, we can use the states feature, to get the location of each state

• This is requred to make the map

states_map <- map_data("state")

• With that, we have all the information for the map

flu_combined <- arrange(merge(states_map, aggregate_data, by.x = "region", by.y = "State"), group, order)
ggplot(flu_combined, aes(x = long, y = lat, group = group, fill = CasesPerPerson)) + geom_polygon(colour = "black") + coord_map("polyconic") + scale_fill_gradient2(low = "lightblue", high = "darkblue", midpoint = median(flu_combined$CasesPerPerson))

• The code goes off the screen, so here’s a better look

flu_combined <- arrange(merge(states_map, aggregate_data, 
                              by.x = "region", by.y = "State"), 
                              group, order)
ggplot(flu_combined, aes(x = long, y = lat, group = group, 
                         fill =CasesPerPerson)) + 
                         geom_polygon(colour = "black") + 
                         coord_map("polyconic") + 
                         scale_fill_gradient2(low = "lightblue", 
                                              high = "darkblue", 
                         midpoint = median(flu_combined$CasesPerPerson))

• Pretty elegant, huh

• As we can see, West Virginia and Mississippi have the highest mortality rate

• I’d imagine some form of age demographics come into play

• On the other hand, Arizona is doing pretty well for itself

• More reason to think I won’t perish and turn into dust

• I know this is slightly scuffed but again, I’m really sick

• I hope you enjoyed