mclosest: Extract closest values in a pairwise manner between two...

In rformassspectrometry/MetaboCoreUtils: Core Utils for Metabolomics Data

Extract closest values in a pairwise manner between two matrices

Description

The mclosest function calculates the closest rows between two matrices (or data frames) considering pairwise differences between values in columns of x and table. It returns the index of the closest row in table for each row in x.

Usage

mclosest(x, table, ppm = 0, tolerance = Inf)

Arguments

x	numeric matrix or data frame representing the query data. Each row in x will be compared to every row in table. Both x and table are expected to have the same number of columns, and the columns are expected to be in the same order.
table	numeric matrix or data frame containing the reference data to be matched with each row of x. Each row in table will be compared to every row in x. Both table and x are expected to have the same number of columns, and the columns are expected to be in the same order.
ppm	numeric representing a relative, value-specific parts-per-million (PPM) tolerance that is added to tolerance (default is 0).
tolerance	numeric accepted tolerance. Defaults to tolerance = Inf, thus for each row in x the closest row in table is reported, regardless of the magnitude of the (absolute) difference.

Details

If, for a row of x, two rows of table are closest only the index of first row will be returned.

For both the tolerance and ppm arguments, if their length is different to the number of columns of x and table, the input argument will be replicated to match it.

Value

integer vector of indices indicating the closest row of table for each row of x. If no suitable match is found for a row in x based on the specified tolerance and ppm, the corresponding index is set to NA.

Author(s)

Philippine Louail

Examples

x <- data.frame(a = 1:5, b = 3:7)
table <- data.frame(c = c(11, 23, 3, 5, 1), d = c(32:35, 45))

## Get for each row of `x` the index of the row in `table` with the smallest
## difference of values (per column)
mclosest(x, table)

## If the absolute difference is larger than `tolerance`, return `NA`. Note
## that the tolerance value of `25` is used for difference for each pairwise
## column in `x` and `table`.
mclosest(x, table, tolerance = 25)

rformassspectrometry/MetaboCoreUtils documentation built on Oct. 2, 2024, 5:13 a.m.