BindSPE: Bind multiple sample data together

In kieranrcampbell/SpatialStats: Statistics for spatially resolved proteomics data

Description

The low power and high noise in some analyses makes it necessary to bind multiple samples together. Given an SPExp object, this method separates out each cell type, calculate neighbour means and binds the samples together, i.e. the individual response matrices for different samples Y1, Y2, Y3 are row-bound together to form one large response matrix Y = [Y1' Y2' Y3']'.

Usage

BindSPE(SPE, choose.class = NULL, use.weights = FALSE, normalize = TRUE)

Arguments

SPE	The SPExp object to use
choose.class	Choose a particular class of cells. Should be a vector of length length{SPE} where the ith entry indicates the class of cells in sample i to pick.
use.weights	Passed to neighbourMeans to optionally weight the regression by relative boundary size
normalize	If true predictor columns are centre-scaled

Details

If normalize = TRUE it is necessary to introduce sample factors as constructed by ConstructSampleFactors, otherwise the regression estimates will be affected by the relative means of different samples.

Value

A list with three components:

• X The bound predictor matrix

• Y The bound response matrix

• sizes A vector with the number of cells selected from each SPE

kieranrcampbell/SpatialStats documentation built on May 20, 2019, 9:24 a.m.