tidyverse.Rmd

---
title: "Tidyverse"
author: "Nnenna Asidianya"
date: "9/25/2021"
output: html_document
---

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

## ggplot2


```{r}
library(tidyverse)
CARS = mtcars
attach(CARS)

#check the scatterplot for mpg against weight

ggplot(CARS, aes(x=wt, y=mpg, color=as.factor(cyl)))+geom_point()+scale_color_manual(values=c("#999999", "#E69F00", "#56B4E9"))


#base R

boxplot(mpg~cyl, col=topo.colors((3)))

#ggplot

ggplot(CARS, aes(x=as.factor(cyl), y=mpg,color=as.factor(cyl)))+geom_boxplot()+scale_color_manual(values=c("#999999", "#E69F00", "#56B4E9"))
```

Note: If I am interested in the characteristics of just one variable. 

```{r}
ggplot(CARS, aes(x=wt))+geom_histogram(bins=5, col="black", fill="red")
```


## readr

Notice in example below the data frames are seemingly read into the global environment in a similar manner.This is in part because the CSV file contains headers for the categories so base R did not need to create a new row.

```{r}
#tidyverse
ages = read_csv("ages.csv")
attach(ages)

#baseR
ages2 = read.csv("ages.csv")
```

## tidyr

This is the modified CSV where the variable labels are removed.

```{r}
AGE = read_csv("AGE.csv", col_names = F)

AGE2 = read.csv("AGE.csv", header = F)
```

We often need to switch between wide and long data format. The `ages_wide` tibble is currently in wide format. To get it in long format we can use `pivot_longer`

```{r}

ages_wide<-ages %>% pivot_wider(names_from = category, values_from = Active:Deaths)

#view(ages_wide)

age_long<-ages_wide %>% gather( Var, Val, -Date) 
```


## dplyr 

```{r}

#in base R
category = fct_relevel(category, "Under 20", "20-29", "30-39", 
                            "40-49", "50-59", "60-69", "70-79","80-89",
                            "90-99", "Unknown")

ages_update  = data.frame(ages, category)



df<-ages %>% mutate(category = fct_relevel(category, "Under 20", "20-29", "30-39", 
                            "40-49", "50-59", "60-69", "70-79","80-89",
                            "90-99", "Unknown")) %>%  filter(!category=="Unknown") %>% 
  group_by(Date) %>% 
  mutate(total_cases=Active+Resolved) %>% 
  mutate(pct_cases=100*(total_cases/sum(total_cases)))

p1<-ggplot(data = df, aes(x = category, y = pct_cases , fill =Date)) +
  geom_bar(stat = "identity", position = position_dodge(), col="black", alpha = 0.75)+
  labs(x = "Ages in Ontario per Decade", y = "Percent", title = "COVID Cases Comparison: Nov. 2020 vs. Feb, 2021 ")

p1<-p1 +  scale_fill_manual("legend", values = c("20-Nov" = "orange", "21-Feb" = "blue"))


```