collapseDim: Collapse dataset dimensions

In pik-piam/magclass: Data Class and Tools for Handling Spatial-Temporal Data

Collapse dataset dimensions

Description

This function will remove names in the data dimension which are the same for each element (meaning that this data dimension contains exactly one element) or, if forced, remove any other subdimension. It is a generalized version of the function collapseNames

Usage

collapseDim(x, dim = NULL, keepdim = NULL)

Arguments

x	MAgPIE object
dim	Either NULL, dimension code or name of dimension or a vector of these. If set to NULL all single entry subdimensions will be removed as they are irrelevant to uniquely identfy a data element. If specified, only the specified subdimensions will be removed (See dimCode for more details how to specify a subdimension). CAUTION: The function also allows to specify subdimensions which are otherwise needed to clearly identify an entry. By removing these subdimensions duplicates in the data will be created potentially causing problems in the further use of the data set. Be careful in removing subdimensions.
keepdim	(only considered if dim is not specified) Can be used to converse single element subdimension which otherwise would get deleted. If dim is specified this setting will not have any effect.

Value

The provided MAgPIE object with collapsed dimensions

Note

This function has some similarities to dimReduce, but serves a different purpose. While collapseDim only removes dimensions which contain only a single element or which it is specifically told to remove, dimReduce looks whether the entries of a multi-entry dimension are all the same and removes dimensions for which this is the case. In some cases both will lead to the same result but in many other cases the results will differ.

Author(s)

Jan Philipp Dietrich

See Also

getItems "magpie"

Examples

x <- new.magpie(c("GLO.1", "GLO.2"), 2000, c("bla.a", "bla.b"))
collapseDim(x)
collapseDim(x, keepdim = 1:2)
collapseDim(x, dim = 1.1)
collapseDim(x, dim = 3.2)

pik-piam/magclass documentation built on March 25, 2024, 11:07 p.m.