Basic usage of R

class: center, middle

# Basic usage of R - 1

## Introductory Computer Programming

### Deepayan Sarkar

---

# Basics of using R

R is more flexible than a regular calculator

* In fact, R is a full programming language

* Most standard data analysis methods are already implemented

* Can be extended by writing add-on packages

* Thousands of add-on packages are available

---

# Major concepts

* Variables (in the context of programming)

* Data structures needed for data analyis

* Functions (set of instructions for performing a procedure)
---

# Variables

* Variables are symbols that may be associated with different values

* Computations involving variables are done using their current value

```r
x <- 10 # assignment
sqrt(x)
```

```
[1] 3.162278
```

```r
x <- -1
sqrt(x)
```

```
Warning in sqrt(x): NaNs produced
```

```
[1] NaN
```

```r
x <- -1+0i
sqrt(x)
```

```
[1] 0+1i
```
---

# Data structures for data analysis

* Vectors

* Matrices

* Data frames (a spreadsheet-like data set)

* Lists (general collection of objects)

---

# Atomic vectors

* Indexed collection of homogeneous scalars, can be

* Numeric / Integer

* Categorical (factor)

* Character

* Logical (`TRUE` / `FALSE`)
	
* Missing values are allowed, indicated as `NA`

* Elements are indexed starting from 1

* $i$th element of vector `x` can be extracted using `x[i]`

* There are also more sophisticated forms of (vector) indexing

---

# Atomic vectors: examples

```r
month.name # built-in 
```

```
 [1] "January"   "February"  "March"     "April"     "May"       "June"      "July"      "August"   
 [9] "September" "October"   "November"  "December" 
```

```r
x <- rnorm(10)
x
```

```
 [1]  0.83685412 -2.42704018  0.09692961  1.55453603  0.77415311 -0.92299041  0.53968367  0.13350713
 [9] -1.44594243  0.18168865
```
--

```r
str(x) # useful function
```

```
 num [1:10] 0.8369 -2.427 0.0969 1.5545 0.7742 ...
```

```r
str(month.name)
```

```
 chr [1:12] "January" "February" "March" "April" "May" "June" "July" "August" "September" ...
```
---

# Atomic vectors: examples

```r
m <- sample(1:12, 30, rep = TRUE)
m
```

```
 [1]  3  2 11 11  6 10  5 10  8  6  4  5  1  6 12  4  5 10  9  6  7  9  4  8  4  3  1  2  1 10
```

```r
mf <- factor(m, levels = 1:12, labels = month.name)
mf
```

```
 [1] March     February  November  November  June      October   May       October   August   
[10] June      April     May       January   June      December  April     May       October  
[19] September June      July      September April     August    April     March     January  
[28] February  January   October  
12 Levels: January February March April May June July August September October ... December
```

```r
str(m)
```

```
 int [1:30] 3 2 11 11 6 10 5 10 8 6 ...
```

```r
str(mf)
```

```
 Factor w/ 12 levels "January","February",..: 3 2 11 11 6 10 5 10 8 6 ...
```

---

# Atomic vectors

* "Scalars" are just vectors of length 1

```r
str(numeric(2))
```

```
 num [1:2] 0 0
```

```r
str(numeric(1))
```

```
 num 0
```

```r
str(0)
```

```
 num 0
```
---

# Atomic vectors

* Vectors can have length zero

```r
numeric(0)
```

```
numeric(0)
```

```r
logical(0)
```

```
logical(0)
```

---

# Types of indexing

* Indexing refers to extracting subsets of vectors (or other kinds of data)

* R supports several kinds of indexing:

* Indexing by a vector of positive integers

* Indexing by a vector of negative integers
	
	* Indexing by a logical vector

* Indexing by a vector of names

---

# Types of indexing

* The "standard" C-like indexing with a scalar (vector of length 1):

```r
month.name[2] # the first index is 1, not 0
```

```
[1] "February"
```

* The "index" can also be an integer vector

```r
month.name[c(2, 4, 6, 9, 11)]
```

```
[1] "February"  "April"     "June"      "September" "November" 
```

* Elements can be repeated

```r
month.name[c(2, 2, 6, 4, 6, 11)]
```

```
[1] "February" "February" "June"     "April"    "June"     "November"
```

---

# Types of indexing

* "Out-of-bounds" indexing give `NA` (missing)

```r
month.name[13]
```

```
[1] NA
```

```r
seq(1, by = 2, length.out = 8)
```

```
[1]  1  3  5  7  9 11 13 15
```

```r
month.name[seq(1, by = 2, length.out = 8)]
```

```
[1] "January"   "March"     "May"       "July"      "September" "November"  NA          NA         
```

---

# Types of indexing

* Negative integers omit the specified entries

```r
month.name[-2]
```

```
 [1] "January"   "March"     "April"     "May"       "June"      "July"      "August"    "September"
 [9] "October"   "November"  "December" 
```

```r
month.name[-c(2, 4, 6, 9, 11)]
```

```
[1] "January"  "March"    "May"      "July"     "August"   "October"  "December"
```

* Cannot be mixed with positive integers

```r
month.name[c(2, -3)]
```

```
Error in month.name[c(2, -3)]: only 0's may be mixed with negative subscripts
```
---

# Types of indexing

* Zero has a special meaning - doesn't select anything

```r
month.name[0]
```

```
character(0)
```

```r
month.name[integer(0)] ## same as empty index
```

```
character(0)
```

```r
month.name[c(1, 2, 0, 11, 12)]
```

```
[1] "January"  "February" "November" "December"
```

```r
month.name[-c(1, 2, 0, 11, 12)]
```

```
[1] "March"     "April"     "May"       "June"      "July"      "August"    "September" "October"  
```
---

# Types of indexing

* Indexing by logical vector: select `TRUE` elements

```r
month.name[c(TRUE, FALSE, FALSE)] # index replicated
```

```
[1] "January" "April"   "July"    "October"
```

```r
i <- substring(month.name, 1, 1) == "J"
i
```

```
 [1]  TRUE FALSE FALSE FALSE FALSE  TRUE  TRUE FALSE FALSE FALSE FALSE FALSE
```

```r
month.name[i]
```

```
[1] "January" "June"    "July"   
```

---

# Types of indexing

* Common use: extract subset satisfying a certain condition (also called "filtering")

```r
(x <- rnorm(20))
```

```
 [1] -2.851806639  0.711520932 -0.649975983  0.338946381  0.554063362 -1.856386243 -1.605922598
 [8] -0.002954457 -0.670840417 -1.839289724 -1.222433477  0.466193505 -0.408518505  0.060178539
[15]  1.307296324  0.072137782 -4.085164248  1.385663548 -0.406905851 -0.722480560
```

```r
x > 0
```

```
 [1] FALSE  TRUE FALSE  TRUE  TRUE FALSE FALSE FALSE FALSE FALSE FALSE  TRUE FALSE  TRUE  TRUE  TRUE
[17] FALSE  TRUE FALSE FALSE
```

```r
x[x > 0]
```

```
[1] 0.71152093 0.33894638 0.55406336 0.46619351 0.06017854 1.30729632 0.07213778 1.38566355
```

```r
mean(x[x > 0])
```

```
[1] 0.612
```

---

# Types of indexing

* Sometimes logical indexing can be replaced by integer indexing using `which()`

```r
i
```

```
 [1]  TRUE FALSE FALSE FALSE FALSE  TRUE  TRUE FALSE FALSE FALSE FALSE FALSE
```

```r
which(i)
```

```
[1] 1 6 7
```

```r
month.name[ which(i) ]
```

```
[1] "January" "June"    "July"   
```

```r
month.name[ -which(i) ] # same as month.name[ !i ]
```

```
[1] "February"  "March"     "April"     "May"       "August"    "September" "October"   "November" 
[9] "December" 
```
---

# Types of indexing

* But be careful about zero-length indices

```r
which(substring(month.name, 1, 1) == "B")
```

```
integer(0)
```

```r
month.name[ which( substring(month.name, 1, 1) == "B") ]
```

```
character(0)
```

```r
-which(substring(month.name, 1, 1) == "B") 
```

```
integer(0)
```

```r
month.name[ -which( substring(month.name, 1, 1) == "B") ]
```

```
character(0)
```

---

class: center middle

# Questions?