Project_Code.rmd

---
title: "Project_Code"
author: "Team 2"
date: "1/13/2023"
output: html_document
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
```


```{r}
# import libraries 
library(ggplot2)
library(dplyr)

# read the first data
case_count <- read.csv('us-states.csv')
head(case_count)

```
```{r}
alb <- dplyr::filter(case_count, state == "Alabama")
sum(alb$cases)

```

```{r}
case_count <- case_count %>% dplyr::mutate(date = as.Date(date) )

# filter dataset by state and case totals 
state_count <- do.call(data.frame, 
                       aggregate(cases ~ state, case_count, FUN = function(x) c(max(x))))

head(state_count)

```
Exploratory Data Analysis on US-States Dataset


```{r, fig.height= 3 , fig.width= 5}

# Create a bar graph of total case count by each US state

#ggsave("plot.png", width = 5, height = 5) Saves last plot
#as 5’ x 5’ file named "plot.png" in working directory.
#Matches file type to file extension.


ggplot(state_count) +
    geom_bar( aes(x=state, y=cases), stat="identity", fill="red", alpha=0.7) + ggtitle('Total Cases per US State/Territory') + theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1)) + theme(plot.title = element_text(hjust = 0.5))

```

```{r}
library(lubridate)


class(ymd("2021-04-19"))
  
```

```{r}

state_case_totals <- function(region) {
  
    filtered_data <- dplyr::filter(case_count, state == region)
    filtered_data[,"cum_cases"] <- cumsum(filtered_data$cases)
  
    label <- paste('Total Number of Covid Cases over Time in', region)
    
    ggplot(data = filtered_data, aes(x=date, y=cases)) + geom_point(color='red') + ggtitle(label) + xlab("Date") + ylab("Number of Cases") + scale_x_date(breaks = scales::pretty_breaks(n=20)) + theme(axis.text.x = element_text(angle = 90, vjust = 0.5)) + geom_text(aes(x=ymd("2021-04-1"), label="Covid Vaccines Avaliable", y=max(filtered_data$cases) - median(filtered_data$cases)), colour="blue", angle=90) + geom_vline(xintercept= ymd("2021-04-19") ,lwd=0.5,colour="blue",label_value("Covid Vaccines Avaliable")) + theme(plot.title = element_text(hjust = 0.5))
}

```





Type any state name to visualize the number of Total/Cumulative cases over time 

```{r}

# Visualizations for each of the US States and territories 
state_case_totals("New York")

```


Exploratory Data Analysis on Vaccine Datasets

1. data cleaning: read the csvs, contain null values, fix data types

```{r}
# read in vaccination data by vaccine type 
jansen_vaccines <- read.csv("COVID-19_Vaccine_Distribution_Allocations_by_Jurisdiction_-_Janssen.csv")
pfizer_vaccines <- read.csv("COVID-19_Vaccine_Distribution_Allocations_by_Jurisdiction_-_Pfizer.csv")
moderna_vaccines <- read.csv("COVID-19_Vaccine_Distribution_Allocations_by_Jurisdiction_-_Moderna.csv")

# for each vaccine dataset, rename the cols to specify the vaccine type
names(jansen_vaccines)[names(jansen_vaccines) == 'X1st.Dose.Allocations'] <- 'Jansen_1st_Dose_Counts'
names(pfizer_vaccines)[names(pfizer_vaccines) == 'X1st.Dose.Allocations'] <- 'Pfizer_1st_Dose_Counts'
names(pfizer_vaccines)[names(pfizer_vaccines) == 'X2nd.Dose.Allocations'] <- 'Pfizer_2nd_Dose_Counts'
names(moderna_vaccines)[names(moderna_vaccines) == 'X1st.Dose.Allocations'] <- 'Moderna_1st_Dose_Counts'
names(moderna_vaccines)[names(moderna_vaccines) == 'X2nd.Dose.Allocations'] <- 'Moderna_2nd_Dose_Counts'

```


Visualize the total number of new vaccine doeses over time for a particular state

```{r}
# Visualization over time for total vaccine doses for each state

merged_state_df1 <- full_join(pfizer_vaccines, moderna_vaccines, by=c('Week.of.Allocations','Jurisdiction'))

# remove dose 1 for pfizer and moderna columns
merged_state_df1 <- merged_state_df1[-c(3,5)]

merged_state_df2 <- full_join(merged_state_df1, jansen_vaccines, by=c('Week.of.Allocations','Jurisdiction'))

# replace Null values with quantity 0
merged_state_df2[is.na(merged_state_df2)] = 0


# covert columns to numeric
new_Df <- merged_state_df2 %>% mutate(Pfizer_2nd_Dose_Counts = as.numeric(Pfizer_2nd_Dose_Counts), Moderna_2nd_Dose_Counts = 
                                        as.numeric(Moderna_2nd_Dose_Counts), Jansen_1st_Dose_Counts = as.numeric(Jansen_1st_Dose_Counts), Week.of.Allocations = as.Date(Week.of.Allocations, "%m/%d/%Y"))


new_Df <- new_Df %>% arrange((Week.of.Allocations))
View(new_Df)

```



```{r}

vaccinations_by_state <- function(region) {
  
    df_filtered <- dplyr::filter(new_Df, Jurisdiction == region)
    
    
    label <- paste('Vaccination Allocations over Time in', region)
    
    # get total of vaccines by summing over each type of vaccine type
    
    df_filtered <- df_filtered %>%
    mutate(vaccine_sum = Pfizer_2nd_Dose_Counts + Moderna_2nd_Dose_Counts + Jansen_1st_Dose_Counts)
    
    # compute cumulative number of vaccines given
    df_filtered[,"cum_doses"] <- cumsum(df_filtered$vaccine_sum)

    ggplot(data = df_filtered, aes(x = Week.of.Allocations, y = cum_doses)) + geom_point() + geom_line(color = 'blue') + ggtitle(label) + xlab("Week of Allocations") + ylab("Number of Total Doses") + theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1)) + scale_x_date(date_labels = "%Y %b") + theme(plot.title = element_text(hjust = 0.5))

}

```


```{r,fig.height= 3 , fig.width= 5}
vaccinations_by_state("New Jersey")

```
Create Bar Graph to show total vaccinations for each state accross all times

```{r, fig.height= 3 , fig.width= 5}
state_totals <- new_Df %>%
    mutate(vaccine_sum = Pfizer_2nd_Dose_Counts + Moderna_2nd_Dose_Counts + Jansen_1st_Dose_Counts)


# filter dataset by state and case totals 
state_totals <- do.call(data.frame, 
                       aggregate(vaccine_sum ~ Jurisdiction, state_totals, FUN = function(x) c(sum(x))))



# Clean the data by combining all state data over time similar to what we did above with the covid cases data

ggplot(state_totals) +
    geom_bar( aes(x=Jurisdiction, y=vaccine_sum), stat="identity", fill="blue", alpha=0.7) + ggtitle('Total Vaccines Administered per US State/Territory') + theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1)) + theme(plot.title = element_text(hjust = 0.5))

```

```{r, fig.height= 3 , fig.width= 5}

# Create map data for US Cases

library(usmap)

plot_usmap(data = state_count, values = "cases", color = "black") + 
  scale_fill_continuous(low = 'white', high = 'red', name = "cases", label = scales::comma) + 
  theme(legend.position = "right") + ggtitle("Total US Covid Cases") + theme(plot.title = element_text(hjust = 0.5))


```

```{r, fig.height= 3 , fig.width= 5}
# Map plot for covid vaccinations for each US State

# Create map data for US Cases

colnames(state_totals)[1] <- "state"
plot_usmap(data = state_totals, values = "vaccine_sum", color = "black") + 
  scale_fill_continuous(low = 'gray', high = 'blue', name = "Vaccines", label = scales::comma) + 
  theme(legend.position = "right") + ggtitle("Total US Covid Vaccinations") + theme(plot.title = element_text(hjust = 0.5))

```
Analysis Comments
1. Strong relationship between States with high Covid cases with states that received higher total vaccine allocations


```{r}

# compute correlation coef R^2 between the two data

final_df <- merge(state_count, state_totals, by = "state")


final_df

```
Correlation Test

```{r}
covid_cases_Data <- final_df$cases
covid_vacc_Data <- final_df$vaccine_sum

corr_coef <- cor(covid_cases_Data, covid_vacc_Data)
print(corr_coef)
```      
A correlation coef value of 0.977 indicates a strong correlation between the number of covid cases and the number of vaccines allocated to that respective state. This supports the conclusion regarding the governments overall strategy to supply more vaccines to states with higher populations and a large load of covid cases.