Vector-class: Vector objects

Description Accessors Coercion Subsetting Convenience wrappers for common subsetting operations Concatenation Displaying See Also Examples

Description

The Vector virtual class serves as the heart of the S4Vectors package and has over 90 subclasses. It serves a similar role as vector in base R.

The Vector class supports the storage of global and element-wise metadata:

  1. The global metadata annotates the object as a whole: this metadata is accessed via the metadata accessor and is represented as an ordinary list;

  2. The element-wise metadata annotates individual elements of the object: this metadata is accessed via the mcols accessor (mcols stands for metadata columns) and is represented as a DataFrame object with a row for each element and a column for each metadata variable. Note that the element-wise metadata can also be NULL.

To be functional, a class that inherits from Vector must define at least a length and a "[" method.

Accessors

In the following code snippets, x is a Vector object.

length(x): Get the number of elements in x.

lengths(x, use.names=TRUE): Get the length of each of the elements.

Note: The lengths method for Vector objects is currently defined as an alias for elementNROWS (with addition of the use.names argument), so is equivalent to sapply(x, NROW), not to sapply(x, length).

NROW(x): Equivalent to either nrow(x) or length(x), depending on whether x has dimensions (i.e. dim(x) is not NULL) or not (i.e. dim(x) is NULL).

names(x), names(x) <- value: Get or set the names of the elements in the Vector.

rename(x, value, ...): Replace the names of x according to a mapping defined by a named character vector, formed by concatenating value with any arguments in .... The names of the character vector indicate the source names, and the corresponding values the destination names. This also works on a plain old vector.

nlevels(x): Returns the number of factor levels.

mcols(x, use.names=TRUE), mcols(x) <- value: Get or set the metadata columns. If use.names=TRUE and the metadata columns are not NULL, then the names of x are propagated as the row names of the returned DataFrame object. When setting the metadata columns, the supplied value must be NULL or a DataFrame object holding element-wise metadata.

elementMetadata(x, use.names=FALSE), elementMetadata(x) <- value, values(x, use.names=FALSE), values(x) <- value: Alternatives to mcols functions. Their use is discouraged.

Coercion

as(from, "data.frame"), as.data.frame(from): Coerces from, a Vector, to a data.frame by first coercing the Vector to a vector via as.vector. Note that many Vector derivatives do not support as.vector, so this coercion is possible only for certain types.

as.env(x): Constructs an environment object containing the elements of mcols(x).

Subsetting

In the code snippets below, x is a Vector object.

x[i]: When supported, return a new Vector object of the same class as x made of the elements selected by i. i can be missing; an NA-free logical, numeric, or character vector or factor (as ordinary vector or Rle object); or a IntegerRanges object.

x[i, j]: Like the above, but allow the user to conveniently subset the metadata columns thru j.

NOTE TO DEVELOPERS: A Vector subclass with a true 2-D semantic (e.g. SummarizedExperiment) needs to overwrite the "[" method for Vector objects. This means that code intended to operate on an arbitrary Vector derivative x should not use this feature as there is no guarantee that x supports it. For this reason this feature should preferrably be used interactively only.

x[i] <- value: Replacement version of x[i].

Convenience wrappers for common subsetting operations

In the code snippets below, x is a Vector object.

subset(x, subset, select, drop=FALSE, ...): Return a new Vector object made of the subset using logical vector subset, where missing values are taken as FALSE. TODO: Document select, drop, and ....

window(x, start=NA, end=NA, width=NA): Extract the subsequence from x that corresponds to the window defined by start, end, and width. At most 2 of start, end, and width can be set to a non-NA value, which must be a non-negative integer. More precisely:

  • If width is set to NA, then start or end or both can be set to NA. In this case start=NA is equivalent to start=1 and end=NA is equivalent to end=length(x).

  • If width is set to a non-negative integer value, then one of start or end must be set to a non-negative integer value and the other one to NA.

head(x, n=6L): If n is non-negative, returns the first n elements of the Vector object. If n is negative, returns all but the last abs(n) elements of the Vector object.

tail(x, n=6L): If n is non-negative, returns the last n elements of the Vector object. If n is negative, returns all but the first abs(n) elements of the Vector object.

rev(x): Return a new Vector object made of the original elements in the reverse order.

rep(x, times, length.out, each) and rep.int(x, times): Repeats the values in x through one of the following conventions:

  • times: Vector giving the number of times to repeat each element if of length length(x), or to repeat the whole vector if of length 1.

  • length.out: Non-negative integer. The desired length of the output vector.

  • each: Non-negative integer. Each element of x is repeated each times.

Concatenation

In the code snippets below, x is a Vector object.

c(x, ..., ignore.mcols=FALSE): Concatenate x and the Vector objects in ... together. Any object in ... should belong to the same class as x or to one of its subclasses. If not, then an attempt will be made to coerce it with as(object, class(x), strict=FALSE). NULLs are accepted and ignored. The result of the concatenation is an object of the same class as x.

Handling of the metadata columns:

  • If only one of the Vector objects has metadata columns, then the corresponding metadata columns are attached to the other Vector objects and set to NA.

  • When multiple Vector objects have their own metadata columns, the user must ensure that each such DataFrame have identical layouts to each other (same columns defined), in order for the concatenation to be successful, otherwise an error will be thrown.

  • The user can call c(x, ..., ignore.mcols=FALSE) in order to concatenate Vector objects with differing sets of metadata columns, which will result in the concatenated object having NO metadata columns.

IMPORTANT NOTE: Be aware that calling c with named arguments (e.g. c(a=x, b=y)) tends to break method dispatch so please make sure that args is an unnamed list when using do.call(c, args) to concatenate a list of objects together.

append(x, values, after=length(x)): Insert the Vector values onto x at the position given by after. values must have an elementType that extends that of x.

expand.grid(...): Find cartesian product of every vector in ... and return a data.frame, each column of which corresponds to an argument. See expand.grid.

Displaying

[FOR ADVANCED USERS OR DEVELOPERS]

Displaying of a Vector object is controlled by 2 internal helpers, classNameForDisplay and showAsCell.

For most objects classNameForDisplay(x) just returns class(x). However, for some objects it can return the name of a parent class that is more suitable for display because it's simpler and as informative as the real class name. See SimpleList objects (defined in this package) and CompressedList objects (defined in the IRanges package) for examples of objects for which classNameForDisplay returns the name of a parent class.

showAsCell(x) produces a character vector parallel to x (i.e. with one string per vector element in x) that contains compact string representations of each elements in x.

Note that classNameForDisplay and showAsCell are generic functions so developers can implement methods to control how their own Vector extension gets displayed.

See Also

Examples

1
showClass("Vector")  # shows (some of) the known subclasses

Example output

Loading required package: stats4
Loading required package: BiocGenerics
Loading required package: parallel

Attaching package: 'BiocGenerics'

The following objects are masked from 'package:parallel':

    clusterApply, clusterApplyLB, clusterCall, clusterEvalQ,
    clusterExport, clusterMap, parApply, parCapply, parLapply,
    parLapplyLB, parRapply, parSapply, parSapplyLB

The following objects are masked from 'package:stats':

    IQR, mad, sd, var, xtabs

The following objects are masked from 'package:base':

    Filter, Find, Map, Position, Reduce, anyDuplicated, append,
    as.data.frame, basename, cbind, colMeans, colSums, colnames,
    dirname, do.call, duplicated, eval, evalq, get, grep, grepl,
    intersect, is.unsorted, lapply, lengths, mapply, match, mget,
    order, paste, pmax, pmax.int, pmin, pmin.int, rank, rbind,
    rowMeans, rowSums, rownames, sapply, setdiff, sort, table, tapply,
    union, unique, unsplit, which, which.max, which.min


Attaching package: 'S4Vectors'

The following object is masked from 'package:base':

    expand.grid

Virtual Class "Vector" [package "S4Vectors"]

Slots:
                                          
Name:    elementMetadata          metadata
Class: DataTable_OR_NULL              list

Extends: "Annotated"

Known Subclasses: 
Class "Hits", directly
Class "Rle", directly
Class "List", directly
Class "Pairs", directly
Class "SelfHits", by class "Hits", distance 2
Class "SortedByQueryHits", by class "Hits", distance 2
Class "SimpleList", by class "List", distance 2
Class "SortedByQuerySelfHits", by class "Hits", distance 3
Class "SortedByQuerySelfHits", by class "Hits", distance 3
Class "HitsList", by class "List", distance 3
Class "SelfHitsList", by class "List", distance 4
Class "SortedByQueryHitsList", by class "List", distance 4
Class "SortedByQuerySelfHitsList", by class "List", distance 5
Class "FilterRules", by class "SimpleList", distance 3

S4Vectors documentation built on Dec. 11, 2020, 2:02 a.m.