Covid_Project.Rmd

---
title: "MA2 Project"
output: word_document
---

```{r setup, include=FALSE}
knitr::opts_knit$set(root.dir = getwd())
```

## R Markdown

This is an R Markdown document. Markdown is a simple formatting syntax for authoring HTML, PDF, and MS Word documents. For more details on using R Markdown see <http://rmarkdown.rstudio.com>.

When you click the **Knit** button a document will be generated that includes both content as well as the output of any embedded R code chunks within the document. You can embed an R code chunk like this:

```{r, include=TRUE, echo=TRUE}
# c.data = read.csv("data/Covid19-FinalData.csv", header=T)
c.data = read.csv("data/Covid-Final.csv", header=T)

str(c.data)
c.data_eda=c.data

c.data$Year = as.factor(c.data$Year)
c.data$Week = as.factor(c.data$Week)
head(c.data,1000)

```

```{r}
library(dplyr)
library(tidyverse)
library(broom)

# Dont  need since  we dont look at whole years data
# ReturnWeeks <-function(year){
#   if (year==2015){
#     print(year)
#   }
#   else {
#     ReturnWeeks(year-1)
#     print(year)
#   }
# }
# filter(c.data_eda,Year==2015)$Week

state_trend = filter(c.data_eda,Year<2020) %>% group_by(State) %>% do(model = lm(Deaths ~ Year, data = .)) %>% rowwise() %>% tidy(model)

state_trend

death_stdv=filter(c.data_eda,Year<2020) %>% group_by(State) %>% do(as.data.frame(t(sd(.$Deaths)))) 
death_stdv

# filter(state_trend,State=='Alabama')$estimate[2]*2020+filter(state_trend,State=='Alabama')$estimate[1]

state_equations=spread(select(state_trend,State,term,estimate),term,estimate)

# c.data_eda$Estimates = c.data_eda %>% rowwise(.) %>% do(filter(state_equations,State==.$State)$"(Intercept)"+filter(state_equations,State==.$State)$Year*.$Year)

#this takes a little while
est = c.data_eda %>% rowwise(.) %>%
  do(as.data.frame(t(filter(state_equations,State==.$State)$"(Intercept)"+filter(state_equations,State==.$State)$Year*.$Year)))


est$V1

c.data_eda$Estimated=est$V1

c.eda= inner_join(c.data_eda,death_stdv, by="State")

# write.csv(c.eda,'data/CVT.csv')

# filter(select(state_equations,"(Intercept)"),"(Intercept)"==2019)
# 
# filter(state_equations,State=='Alabama')$"(Intercept)"+filter(state_equations,State=='Alabama')$Year*2020
# 
# 
# state_equations$"(Intercept)"+state_equations$Year*2020


```


```{r, include=TRUE, echo=TRUE}
#poisson
#we should instead create a weekly baseline trend indexed by state rather then treat each week as 
#then test for seasonal (my impression is those are small) and see if we can just collapse 
#the deaths to a weekly baseline with std deviation
c.glm_p = glm(Deaths~Week+Covid+PopDensity+Population+Year+TAVG,
data=c.data, family=poisson)
summary(c.glm_p)

print("The AIC of Poisson Log linear Model is  ")
AIC(c.glm_p)
print("The BIC of Poisson Log linear Model is ")
BIC(c.glm_p)
```


```{r, include=TRUE, echo=TRUE}
#Poisson with RE

library(lme4)
m.re1 = glmer(Deaths~Week+Covid+PopDensity+Population+Year+TAVG+(1|State),
data=c.data, family=poisson)
summary(m.re1)

print("The AIC of Poisson Log linear Model with different Intercepts is  ")
AIC(m.re1)
print("The BIC of Poisson Log linear Model with different Intercepts is ")
BIC(m.re1)



```

## Including Plots

You can also embed plots, for example:

```{r pressure, echo=FALSE}
plot(pressure)
```

Note that the `echo = FALSE` parameter was added to the code chunk to prevent printing of the R code that generated the plot.