Meeting2Notes.Rmd

---
title: "Meeting 2: R and Ebola"
author: "Brooke Anderson"
date: "November 14, 2014"
output: ioslides_presentation
---

## Last time, we read in our data

```{r}
## If necessary, use setwd() to get to the right directory
ebola <- read.table("country_timeseries.csv", sep = ",",
                    header = TRUE)
ebola[1:3, 1:5]
```

## Today's plan

- Dataframes and vectors
- Subsetting 
- Functions
- Example of a function-- the plot function

# Dataframes and vectors

## Dataframes and vectors

A vector is a string of values:

Example 1: Start of the vector of the dates when Ebola cases were reported
```{r, echo = FALSE}
as.character(ebola$Date[1:10])
```

Example 2: Start of the vector of number of cases reported in Guinea
```{r, echo = FALSE}
ebola$Cases_Guinea[1:10]
```

## Dataframes and vectors

You can make a new vector using the concatenation vector, `c(...)`:

```{r}
x <- c(1, 5, 7, 9, 10)
x
class.names <- c("Taylor", "Maggie", "Mimi", "Brianna", "Jon")
class.names
```

## Dataframes and vectors

A dataframe is made up of a lot of vectors stuck together
*(Notice how each column is a vector)*

```{r, echo  = FALSE}
ebola[1:10, 1:5]
```

## Dataframes and vectors

You can make a new dataframe using the function `data.frame()`:

```{r}
class.data <- data.frame(name = class.names,
                         number = x)
class.data
```

# Subsetting

## Subsetting

You can use indexing (`[...]`, `[..., ...]`) to subset from a vector or dataframe, like:

```{r, eval=FALSE}
vector[locations]  ## Generic code
dataframe[row locations, column locations] ## Generic code
```

## Subsetting

A vector has one dimension, so you index without a comma (i.e., in one dimension):

```{r}
class.names[1]
class.names[c(2, 3, 4)]  ## Equivalent: class.names[2:4]
```

## Subsetting

A dataframe has two dimensions (rows and columns), so you index with a column:

```{r}
class.data[1,1]
class.data[1:3, 1:2]
```

## Subsetting

To get all values in a dimension (row or column), leave that part of the index blank:

```{r}
class.data[1, ]
class.data[ , 1]
```

## Subsetting

For columns, you can use column names instead of location:

```{r}
class.data[3:4, "number"]
class.data[3:4, c("name", "number")]
```

## Subsetting

You can also pull a column (vector) from a dataframe using `$`, like:

```{r, eval=FALSE}
dataframe$column.name  ## Generic code
```

For example, to get the column of `ebola` with cases from Guinea:
```{r}
head(ebola$Cases_Guinea)
```

*Note: I've used `head` to look at just the start of the vector since the whole thing would be really long.*

## Now you try...

Try to get the following vectors from the dataset:

- Date
- The ten most recent counts of cases in the US
- The earliest twenty counts of deaths in Liberia
- A dataframe of the first five observations of date, cases in Mali and deaths in Mali

*Hint: Try using `colnames(ebola)` to find out the names of all the columns in `ebola`. Also, use `dim(ebola)` to find out the dimensions of the dataframe so you can get the index numbers right for the latest ten data points.*

# Functions

## Functions

In general, functions in R take the following structure:

```{r, eval = FALSE}
function.name(required information, options)  ## Generic code
```

The result of the function will be output to your R session, unless you choose to save the output in an object:

```{r, eval = FALSE}
new.object <- function.name(required information, options)  ## Generic code
```

## Functions

Examples of this structure:

```{r, eval = FALSE}
head(ebola)
head(ebola, n = 3)
ebola <- read.table("country_timeseries.csv", sep = ",",
                    header = TRUE)
```

Find out more about a function by using `?function.name`. This will take you to the help page for the function, where you can find out all the possible arguments for the function, required and optional.

# Example of a function

## The `plot` function

The plot function has two required arguments: the x coordinates of points in the plot, and the y coordinates of points in the plot. The generic structure is:

```{r, eval=FALSE}
plot(x = x coordinates, y = y coordinate)  ## Generic code
```

As long as you put the x coordinates first and the y coordinates second, you can leave out the `x = ` and `y = `:

```{r, eval = FALSE}
plot(x coordinates, y coordinate)  ## Generic code
```

## The `plot` function

```{r, fig.width = 4.5, fig.height = 3.5, fig.align = "center"}
x <- c(1, 2, 3)
y <- c(4, 5, 6)
plot(x, y)
```


## The `plot` function

Now that you know how to pull out two vectors you want from the ebola dataset, you can plot them:

```{r, fig.height = 4, fig.width = 5, fig.align = 'center'}
plot(ebola$Day, ebola$Cases_Guinea)
```

## The `plot` function

The plot function also has *many* optional arguments (check `?plot`, `?plot.default`). For example, 

- `type`: What do you want to plot? Points (`"p"`)? Lines (`"l"`)?
- `main`: Give a title to your plot (`"My title`")
- `xlab`, `ylab`: Give nicer labels to your x- and y-axes (`xlab = "Day in Ebola data collection"`)
- `xlim`, `ylim`: Specify the range of your x- and y- axes (`xlim = c(0, 100)`)

## The `plot` function

```{r, fig.height = 4, fig.width = 5, fig.align = 'center'}
plot(ebola$Day, ebola$Cases_Guinea, main = "Guinea ebola cases",
     xlab = "Day in ebola data collection", ylab = "# of cases",
     type = "l", lwd = 2, col = "gray")
```

## Now you try...

Try plotting:

- Deaths in Liberia by day
- Mortality rate in Liberia by day
- Deaths in Liberia by date

Experiment with options like `type`; `col`, `pch` and `cex` (when you're plotting points, `type = "p"`), `lwd` (when you're plotting lines, `type = "l"`), `main`, `sub`, `xlim`, and `ylim`.

*Hint: Try using `colnames(ebola)` to find out the names of all the columns in `ebola`.*