20190918_rmarkdown_dataviz.Rmd

---
title: "Data Visualization with Rmarkdown"
author: "Ariel Gershman"
date: "9/18/2019"
runtime: shiny
output:
  html_document:
    code_download: true
---

```{r setup, include=F}
# up there is the yaml header, if you launch from Rstudio it makes it for you
knitr::opts_chunk$set(echo = TRUE)
```

```{r libs, include=F}
# load libraries some of these libraries you will need to install with install.packages("library")
library(tidyverse)
library(plotly)
library(maps)
library(shiny)
library(knitr)
library(kableExtra)
```
## R Markdown basics 

We can type plain text for comments and descriptions about the data and analysis

### we can change the size of our text

and make important things __bold__

or set up [links](https://www.rstudio.com/wp-content/uploads/2015/02/rmarkdown-cheatsheet.pdf) to helpful resources 

we can also print equations in the text $A = \pi*r^{2}$ 

and highlight important things in block text 

>Embedding code can be done in two ways
>
>in-line code: two plus two equals `r 2 + 2` 
>
>or in code chunks:

```{r inline, eval=T, include = T, tidy = T}
p <- 2 + 2 
print(p)
```

## Let's get started
```{r storms, include=F}
# first save this data to .RData so it will be found when we knit 
dat <- storms
save(dat, file="dat.RData")
load("dat.RData")
```

### Including Plots

```{r types, echo=F}
# number of each type of storm per year
# first we want to parse the data using tidyverse then pipe it into ggplot for visualization
# echo = FALSE because we don't want this code printed in our doc but we want the output to be printed 
p1 <- dat %>% # load data 
  group_by(year, status) %>% # pipe data into grouping by year and status
  summarise(number = n()) %>% # count the number of storms of each status per year 
  ggplot(aes(x = year, y  = number, fill = status))+theme_classic()+labs() +
  geom_bar(stat = "identity", position = "stack") + labs(x = "year", y = "number of storms") # plot a bargraph of the number of storms per year making each status and different color 

(gg <- ggplotly(p1)) # use the plotly package to make the graph interactive
```

```{r speed, echo=F}
p2 <- dat %>%
  unite("Date", c(year,month,day), sep = "-") %>% # format the date so we can order the data by date
  mutate(Date = as.Date(Date)) %>%
  group_by(Date,status) %>% # group by the date and the statys and calculate mean wind spees per day for each storm type
  summarise(wind = mean(wind)) %>%
  ggplot(aes(x = wind))+geom_histogram(aes(fill = status), alpha = .4, bins = 30)+ theme_classic()+labs(x = "wind speed", y = "count") # plot a histogram of the average wind speed per day and color by storm type, alpha is the opacity 
(gg <- ggplotly(p2)) # make interactive with plotly 
```

##Let's make some summary stats

```{r sum, include=FALSE}
# include = FALSE because we don't want any of this printed in our doc 
sum <- dat %>%
  group_by(status) %>%
  summarise(wind = mean(wind), pressure = mean(pressure))
```

###Mean Wind Speed:

Hurricane: `r sum$wind[1]` mph

Tropical Depression: `r sum$wind[2]` mph

Tropical Storm: `r sum$wind[3]` mph

## Including Tables
```{r echo = F, results="asis"}
# the kable package is an easy way to make nice looking tables 
kable(dat[1:20,1:11], caption = "a knitr table") %>%
  kable_styling(bootstrap_options = "striped", fixed_thead = T,font_size = 18) %>%
   scroll_box(width = "700px", height = "200px")
```

## Reactive plotting
```{r slider, echo=F}

# we can use the shiny package to make interactive plots 
# first we set the input options 
inputPanel(
 selectInput("year", label = "Year:",
choices = dat$year),
sliderInput("wind", label = "Wind Speed:", min = min(dat$wind), max = max(dat$wind), value = min(dat$wind))
)
# then plot with the variables that the user inputs 
renderPlot({
  subdat <- subset(dat, year == input$year & wind >= input$wind)
ggplot() +
  geom_polygon(data = map_data("world"),
  aes(x = long, y = lat, group = group)) + geom_point(data = subdat, aes(x = long, y = lat, color = name))+theme_bw()
 })


```