Reshaping

The goal is to get tidy data. That means to have:

• Each variable forms a column

• Each observation forms a row

• Each table/file stores data about one kind of observations.

Data to reshape

We will use standard R data set mtcars.

head(mtcars)

Install reshape2 package

Nice article about respahe2

install.packages("reshape2")
library(reshape2)

Melting data set

We will pass the following variables into melt function:

• data set

• what columns are IDs

• what columns are variables

The following code will add one more column with row names and primarilly, make IDs from "carname", "gear" and "cyl". The remaining columns will be melted down.

mtcars$carname <- rownames(mtcars)

carMelt <- melt(
    mtcars,
    id=c("carname", "gear", "cyl"),
    measure.vars=c("mpg", "hp")
)

Let's have a look at the melted values.

> head(carMelt, n=3)
        carname gear cyl variable value
1     Mazda RX4    4   6      mpg  21.0
2 Mazda RX4 Wag    4   6      mpg  21.0
3    Datsun 710    4   4      mpg  22.8
...
62  Ferrari Dino    5   6       hp   175
63 Maserati Bora    5   8       hp   335
64    Volvo 142E    4   4       hp   109

The better way to see what happend is to display values in table. Watch values in variable column.

View(carMelt)

Casting data sets

We can cast the data set into different shapes.

> cylData <- dcast(carMelt, cyl ~ variable)
Aggregation function missing: defaulting to length

> cylData

  cyl mpg hp
1   4  11 11
2   6   7  7
3   8  14 14

We can specify function to aggregate values.

> cylData <- dcast(carMelt, cyl ~ variable, mean)

> cylData

  cyl      mpg        hp
1   4 26.66364  82.63636
2   6 19.74286 122.28571
3   8 15.10000 209.21429

Average Values

We will InsectSprays data set.

> head(InsectSprays)

  count spray
1    10     A
2     7     A
3    20     A
4    14     A
5    14     A
6    12     A

The following code is using tapply function to show sum of count column for each spray.

> tapply(InsectSprays$count, InsectSprays$spray, sum)

  A   B   C   D   E   F
174 184  25  59  42 200

Split values

> split <- split(InsectSprays$count, InsectSprays$spray)

> split
$A
 [1] 10  7 20 14 14 12 10 23 17 20 14 13

$B
 [1] 11 17 21 11 16 14 17 17 19 21  7 13
...

Split and sum

> splitSum <- lapply(split, sum)

> splitSum
$A
[1] 174

$B
[1] 184
...

Combine

We can combine the values craeted above in two ways.

unlist(splitSum)

sapply(split, sum)

Using plyr package to summarize values

Be careful, there is summarise and NOT summarize. If you do that mistake you get argument "by" is missing, with no default error.

ddply(InsectSprays, .(spray), summarise, sum=sum(count))

More to lear on http://www.r-bloggers.com/a-fast-intro-to-plyr-for-r

Other functions

• acast casting multi-dimensional array

• arrange faster reordering without using order() function

• mutate adding new variables