Package 'morphology'

List with the distance matrix and within, and potentially a connected matrix, typically output from look_for(). Alternatively, this can be a data frame resulting from a describe() including a name argument, to add additional described summaries.

Summary function.

Quantity to summarise, and return as y. Can be "neighbours" (default), "distance" or "pathlength".

Quantity to return as x. The default is to use "distance" in case of is "neighbours" or "pathlength", and "neighbours" otherwise.

Distances at which to describe the results. Defaults to the integer sequence from 0L to the maximum search distance, if available, and otherwise the maximum distance found.

Logical. Count connected or disconnected neighbours only. Defaults to NULL, for which all neighbours are considered.

Logical. Provide cumulative results (distances <= threshold) or binned results (distances at center of provided breaks; default).

Name to use for the current described dataset in an extra name column. If provided, the input list is included as attribute to the output, and describe() can be repeated multiple times in a pipechain.

Object to remove the attributes from.

Dimensions of the generated array.

Proportion of the spheres category in the generated array.

Radius of the spheres.

Array to analyze.

Either "voxels" (default) or "components" to look at individual voxels or connected components.

Vector of categories to consider. If negative, these are omitted.

Vector of categories to consider for cross-category morphology. If negative, these are omitted.

Width of the kernel for determining connectivity. Can be a vector (row, column, layer). Not used when looking at voxels.

Shape of the kernel for determining connectivity. Can be "diamond" (default), "disc" or "box". Not used when looking at voxels.

Data frame with x, y, and/or z, and value or component columns, typically output from look_at() or look_in().

Number of nearest neighbours to look for. Can be Inf if within is finite, in which case all neighbours within that distance are looked for.

Maximum search distance.

Look for neighbours of border voxels. Defaults to TRUE. If FALSE, minus sampling is performed.

Approximate error bound, for approximate nearest neighbour search.

Data frame with x, y, z and value or component columns, typically output from look_at().

Direction to consider. Either 1D ("x", "y", "z"), 2D ("xy", "xz", "yz") or 3D ("xyz").

Value indicating how many voxels to consider, e.g. one of two (1/2), one every three (1/3), etc. Defaults to every voxel (1/1). This can be a vector (row, column, layer) to thin the different dimensions differently. Alternatively it can be a list of two vectors, where the first one is used for the category of interest (of_category in look_at()) and the second one is for the neighbour category (in_relation_to in look_at(), or again of_category for a self-query).

Arguments passed to the quantile() function.

Arguments passed to the quantile() function.

Arguments passed to the quantile() function.

Arguments passed to the quantile() function.

Arguments passed to the quantile() function.

Data frame

Things to scale the description by. Can be "neighbourhood", which automatically chooses between alternative options "volume", "area" or "length", or "proportion", "inverse proportion", another description data frame with the same distances, ...

Data frame resulting from a describe() call, or just a numeric vector.

Arguments passed to ggplot2::labs() for labelling.