superset.character {nonmemica}

Coerce to Superset from Character

Description

Coerces to superset from character, treating x as a model name.

Usage

## S3 method for class 'character'
superset(
  x,
  read.input = list(read.csv, header = TRUE, as.is = TRUE),
  read.output = list(read.table, header = TRUE, as.is = TRUE, skip = 1, comment.char =
    "", check.names = FALSE, na.strings = c("", "\\s", ".", "NA")),
  include = character(0),
  exclude = character(0),
  rename = NULL,
  digits = 5,
  visible = "VISIBLE",
  after = NULL,
  groups = character(0),
  imputation = generalize,
  ...
)

Arguments

x

object

read.input

a methodology for acquiring the input

read.output

a methodology for acquiring the output

include

column names in output to consider adding

exclude

column names in output to reject

rename

logical: whether to keep and rename columns with re-used names

digits

significant digits for assessing informativeness when exclusive=NULL

visible

a name for the flag column indicating visibility

after

place new columns after this column; at end by default (NULL); TRUE places them after last model-visible column (see input statement)

groups

character vector of groupings within which any imputations will be performed

imputation

a list of functions (or arguments to match.fun()) to perform imputations within cells defined by groups: e.g. generalize, forbak, etc (to be tried in succession for new columns only).

...

passed arguments

Details

Given a model name, (project passed or set as global option) superset() figures out the run directory and location of a NONMEM control stream. It reads the control stream to identify the run-time location of input and output files, as well as the "ignore" (and/or "accept") criteria that relate extent of input records to extent of output records. 'read.input' and 'read.output' are lists consisting of functions and arguments appropriate for reading input and output file formats, respectively. The ignore criteria will be reconstructed per row so that output can be mapped unambiguously to input. A column named VISIBLE is bound to the input data, showing 1 where a record was visible to NONMEM, and 0 otherwise. During integration, naming convention of the input is retained, and output column names are mapped by position, using the control stream input criteria. Output tables are restored to input dimensions using the "ignore" criteria, then checked for length: currently, superset ignores output tables having fewer rows than the input, as well as output tables whose row count is not a multiple of input row count. Output tables may contain versions of input columns. Disposition depends on the values of include, exclude, and rename. If include has length, other columns are excluded. Then, if exclude has length, these columns are excluded. Then, if rename is FALSE all remaining columns with re-used names will be dropped. If TRUE, such columns will be renamed (*.n, where n is table number). If NULL, only informative columns will be retained and renamed. A column is informative if any element is informative. An element is informative if it is newly generated (not NA and not zero, but original is NA) or if it is an alteration (non-NA, and different from non-NA original). If the column pair can be interpreted as numeric, "different" is determined using only the first digits digits. Only the first instance of any column among successive output tables is retained. In the control stream, avoid use of FIRSTONLY, as this alters the number of rows.

Value

superset: a data.frame where row count is a multiple of (typically equal to) input row count.

See Also

Other superset: generalize(), ignored(), meta.character(), meta.numeric(), metaplot.character(), metaplot.numeric(), metaplot_character(), metasuperset(), meta(), ninput.character(), ninput.numeric(), ninput(), shuffle(), superset.numeric(), superset(), superspec.character(), superspec.numeric(), superspec()

Examples

library(magrittr)
library(dplyr)
library(wrangle)
options(project = system.file('project/model',package='nonmemica'))
1001 %>% superset %>% head
1001 %>% superset %>% filter(VISIBLE == 1) %>% group_by(ID,TIME) %>% status
1001 %>% ignored %>% table