WGCNA (version 1.73)

pquantile: Parallel quantile, median, mean

Description

Calculation of ``parallel'' quantiles, minima, maxima, medians, and means, across given arguments or across lists

Usage

pquantile(prob, ...)
pquantile.fromList(dataList, prob)
pmedian(...)
pmean(..., weights = NULL)
pmean.fromList(dataList, weights = NULL)
pminWhich.fromList(dataList)

Value

pquantile, pquantile.fromList

A vector or array containing quantiles.

pmean, pmean.fromList

A vector or array containing means.

pmedian

A vector or array containing medians.

pminWhich.fromList

A list with two components: min gives the minima, which gives the indices of the elements that are the minima.

Dimensions are copied from dimensions of the input arguments. If any of the input variables have dimnames, the first non-NULL dimnames are copied into the output.

Arguments

prob

A single probability at which to calculate the quantile. See quantile.

dataList

A list of numeric vectors or arrays, all of the same length and dimensions, over which to calculate ``parallel'' quantiles.

weights

Optional vector of the same length as dataList, giving the weights to be used in the weighted mean. If not given, unit weights will be used.

...

Numeric arguments. All arguments must have the same dimensions. See details.

Author

Peter Langfelder and Steve Horvath

Details

Given numeric arguments, say x,y,z, of equal dimensions (and length), the pquantile calculates and returns the quantile of the first components of x,y,z, then the second components, etc. Similarly, pmedian and pmean calculate the median and mean, respectively. The funtion pquantile.fromList is identical to pquantile except that the argument dataList replaces the ... in holding the numeric vectors over which to calculate the quantiles.

See Also

quantile, median, mean for the underlying statistics.

Examples

# Generate 2 simple matrices
a = matrix(c(1:12), 3, 4);
b = a+ 1;
c = a + 2;

# Set the colnames on matrix a

colnames(a) = spaste("col_", c(1:4));

# Example use

pquantile(prob = 0.5, a, b, c)

pmean(a,b,c)
pmedian(a,b,c)