RleDataFrame-class: Class '"RleDataFrame"'

In genoset: A RangedSummarizedExperiment with methods for copy number analysis

Description

The RleDataFrame class serves to hold a collection of Run Length Encoded vectors (Rle objects) of the same length. For example, it could be used to hold information along the genome for a number of samples, such as sequencing coverage, DNA copy number, or GC content. This class inherits from both DataFrame and SimpleRleList (one of the AtomicVector types). This means that all of the usual subsetting and applying functions will work. Also, the AtomicList functions, like mean and sum, that automatically apply over the list elements will work. The scalar mathematical AtomicList methods can make this class behave much like a matrix (see Examples).

New objects can be created with the RleDataFrame constructor: RleDataFrame(..., row.names=NULL), where ... can be a list of Rle objects, or one or more individual Rle objects.

Use in Biobase eSet objects

The genoset class defines an annotatedDataFrameFrom method for DataFrame, which makes it possible to include DataFrames as assayData elements. The column names for DataFrame cannot be NULL, which makes it impossible to use them as assays in SummarizedExperiment at this time.

Row and Column Summaries

These objects will sometimes be in place of a matrix, as in the eSet example above. It is convenient to have some of the summarization methods for matrices. Each of these methods takes an RleDataFrame and returns a single Rle. The time required is similar to that required for a matrix. For an RleDataFrame x,

rowSums:

Sum across 'rows'.

rowMeans:

Means across 'rows'.

colSums:

Sum each Rle. This is just the sum method for SimpleRleList.

colSums:

Mean of each Rle. This is just the mean method for SimpleRleList.

Slots

rownames:

Object of class "character_OR_NULL" Names to describe each row of the DataFrame. These may end up taking more space than your collection of Rle objects, so consider leaving this NULL.

nrows:

Object of class "integer" Number of rows.

elementType:

Object of class "character" Notes that elements of the internal list are Rle objects.

elementMetadata:

Object of class "DataFrame_OR_NULL" Metadata on the elements, see DataFrame.

metadata:

Object of class "list" Metadata on the whole object, see DataFrame.

listData:

Object of class "list" Base list containing the Rle objects.

Extends

Class "SimpleRleList", directly. Class "DataFrame", directly.

Methods

as.matrix

signature(x = "RleDataFrame"): Convert to matrix.

coerce

signature(x = "RleDataFrame"): Convert to other classes.

colMeans

signature(x = "RleDataFrame"): Mean of each column.

colSums

signature(x = "RleDataFrame"): Sum of each column.

rowMeans

signature(x = "RleDataFrame"): Mean of each 'row'.

rowSums

signature(x = "RleDataFrame"): Sum of each 'row'.

show

signature(object = "RleDataFrame"): Short and pretty description of an object of this type.

Author(s)

Peter M. Haverty, design suggestion from Michael Lawrence.

See Also

DataFrame AtomicList Rle RleList rowMeans colMeans rowSums colSums view-summarization-methods

Examples

showClass("RleDataFrame")

## Constructors
df = new("RleDataFrame", listData=list(A=Rle(c(NA, 2:3, NA, 5), rep(2,
5)), B=Rle(c(6:7, NA, 8:10),c(3,2,1,2,1,1))), nrows=10L)

df2 = RleDataFrame(list(A=Rle(c(NA, 2:3, NA, 5), rep(2, 5)),
B=Rle(c(6:7, NA, 8:10),c(3,2,1,2,1,1))))

df3 = RleDataFrame(A=Rle(c(NA, 2:3, NA, 5), rep(2, 5)), B=Rle(c(6:7,
NA, 8:10),c(3,2,1,2,1,1)))

## AtomicList Methods
runValue(df)
runLength(df)
ranges(df)
mean(df)
sum(df)
df + 5
log2(df) - 1

## Row and Column Summaries
rowSums(df)
colSums(df)
rowMeans(df)
colMeans(df)

## Coercion
as(df, "matrix")
as(df, "list")
as(df, "RleList")
as(df, "DataFrame")
as(df, "data.frame")

genoset documentation built on Nov. 8, 2020, 6:07 p.m.