chunk: split IDs into groups to use for subsequent plotting

In metrumresearchgroup/PKPDmisc: Pharmacokinetic and Pharmacodynamic Data Management Functions

split IDs into groups to use for subsequent plotting

Description

split IDs into groups to use for subsequent plotting

chunk

Usage

ids_per_plot(id, id_per_plot = 9)

chunk(.x, .nchunk = parallel::detectCores())

chunk_grp(.x, .nchunk = parallel::detectCores())

chunk_list(.x, .nchunk = parallel::detectCores())

chunk_grp_list(.x, .nchunk = parallel::detectCores())

Arguments

id	vector of ids (eg id column)
id_per_plot	number of ids per plot. Default to 9
.x	vector of values
.nchunk	number of chunks to identify

Details

works very well with hadley wickham's purrr package to create a column to split on then subsequently plot, see vignette("Multiplot") for details

chunk by group, unique values, and return as a vector or a list with elememts

Examples

#chunking will provide the chunk index by splitting the data as evenly as possible
# into the number chunks specified
letters[1:9]

chunk(letters[1:9], 3)

letters[c(1, 1, 1:7)]
chunk(letters[c(1, 1, 1:7)], 3)

# sometimes you want to evenly chunk by unique values rather than purely balancing
chunk_grp(c(1, 1, 1:7), 3)

# a next step after chunking is splitting into a list, so this does thus for you

# chunk list will both split the data and keep the original values
chunk_list(letters[1:9], 3)

chunk_list(c(letters[1], letters[1], letters[1:7]), 3)

# in this case ragged arrays will be created to keep the number of 
# unique elements consistent as possible between chunks
chunk_grp_list(c(letters[1], letters[1], letters[1:7]), 3)

metrumresearchgroup/PKPDmisc documentation built on Oct. 21, 2022, 9:41 a.m.