Lecture 4 Subsetting.Rmd

---
title: 'Lecture 4: Subsetting vectors'
author: "David McAllister"
date: "Monday 31st October, 2016"
output:
  ioslides_presentation: default
---

## Objective of section
* To be able to use different types of indexing ([, [[ and $)
* To be able to use integer, negative integer and logical values within these

## Much of the following is taken from
Advanced R by Hadley Wickham
"http://adv-r.had.co.nz/"

## The little operator that could `[` 
R does all of this

* Filtering

* Deleting

* Changing values

* Merging

## Why subsetting is hard

* three subsetting operators, [, [[ and $

* four types of subsetting, +ve integer, -ve integer, logical and character

* Different for different objects (e.g. vectors,lists, factors, matrices and data frames)

* The dreaded drop, subsetting changes object types

## Atomic vectors

```{r}
x <- c(2.1, 4.2, 3.3, 5.4)
```

The number after the decimal point gives the original position in the vector.

## __Positive integers__ return elements at the specified positions:

```{r}
x[c(3, 1)]
x[order(x)]

# Duplicated indices yield duplicated values
x[c(1, 1)]

# Real numbers are silently truncated to integers
x[c(2.1, 2.9)]
```

## __Negative integers__ omit elements at the specified positions:

```{r}
x[-c(3, 1)]
```

    You can't mix positive and negative integers in a single subset:

```{r, error = TRUE}
x[c(-1, 2)]
```

## __Logical vectors__ 
select elements where the corresponding logical value is `TRUE`
Most useful type of subsetting

```{r}
x[c(TRUE, TRUE, FALSE, FALSE)]
x > 3
x[x > 3]
```

## Logical versus integer vectors
For logical vector, length of indexing vector = length of vector

```{r}
x[c(TRUE, FALSE)]
# Equivalent to due to vector recycling
x[c(TRUE, FALSE, TRUE, FALSE)]
```

## Missing indexing
Missing value in indexing vector yields missing value in output

```{r}
x[c(TRUE, TRUE, NA, FALSE)]
```

## __Nothing__ returns the original vector.

```{r}
x[]
```


## __Character vectors__ 
Used to return elements with matching names.
```{r}
(y <- setNames(x, letters[1:4]))
y[c("d", "c", "a")]

# Like integer indices, you can repeat indices
y[c("a", "a", "a")]

# When subsetting with [ names are always matched exactly
z <- c(abc = 1, def = 2)
z[c("a", "d")]
```

# Assigning

## Typical data cleaning exercise

```{r, echo = c(-1,-2,-3, -10)}
x <- round(rnorm(10, mean = 120, sd = 20), 0)
x[c(1,4,8)] <- -99
x
# Integer subsetting with assignment
x[1] <- NA
# Logical subsetting with assignment
x[x == -99] <- NA
x
```

## Typical data modifying exercise
```{r, echo = -5}
# More complex logical subsetting with assignment
sex <- rep(c("Male", "Female"), each = 5) 
x_mod <- x
x_mod [sex == "Male"] <- x_mod[sex == "Male"] * 1.2
data.frame (sex = sex, x = x, x_mod = x_mod)
```