Extract the parameter results from a PKNCAresults and return them as a data.frame.

Usage

# S3 method for class 'PKNCAresults'
as.data.frame(
  x,
  ...,
  out_format = c("long", "wide", "cdisc"),
  filter_requested = FALSE,
  filter_excluded = FALSE,
  grpid_cols = NULL,
  grpid_numeric = NULL,
  out.format = deprecated()
)

Arguments

x

The object to extract results from

...

Ignored (for compatibility with generic as.data.frame())

out_format

Should the output be 'long' (default), 'wide', or 'cdisc'? When 'cdisc', the original PKNCA parameter name is kept in a new pknca_parameter column (lowercase so it cannot be mistaken for an SDTM PP variable; drop it before submission), the PPTESTCD column is translated to CDISC standard codes, and a PPTEST column with the CDISC test name is added. The translation is many-to-one – several PKNCA parameters can resolve to the same PPTESTCD (every AUCint variant to "AUCINT", for example) – so pknca_parameter is the only column that still identifies which PKNCA calculation produced a row. Route-dependent parameters (e.g. CL, VZ, MRT) are resolved using the route information from the dose data.

Each row also gets its time point reference, as SDTMIG 3.4 defines it: PPSTINT and PPENINT give the interval start and end as ISO 8601 durations relative to the reference named in PPTPTREF, and PPRFTDTC gives that reference's date-time. The reference is the dose that starts the interval: the last included dose at or before the interval start for that subject and group (PPTPTREF "LAST DOSE PRIOR TO INTERVAL"), so a steady-state dosing interval reads "PT0H" to "PT24H". A subject (or group) without an included dose uses its first concentration instead (PPTPTREF "FIRST OBSERVATION"), as the date-time references of PKNCAdata() do. The durations are in the preferred time unit (timeu_pref) when it is given; PPENINT is NA for an interval to infinity, and PPSTINT, PPENINT, and PPTPTREF are NA when the subject has doses but none at or before the interval start. PPRFTDTC is only given when the concentration and dose times were date-times (see PKNCAdata()), and it can differ between the intervals of one subject.

PPGRPID, the record group identifier, is built from the grouping columns named in grpid_cols and the number of the row's interval, as "<prefix1><value1>.<prefix2><value2>.I<nn>" (for example "A.P1.I01", "P2.I01", or just "I01" without grpid_cols). Within each combination of the subject, the analyte, and the grpid_cols columns, the intervals are numbered from 1 by start and then end; the rows of one interval (one per parameter) share the number. Other grouping columns (a treatment or a matrix, for example) do not enter PPGRPID: they stay as their own output columns, and intervals with the same window in different values of them share a number. The number is written with at least two digits, and with as many as the largest interval number needs, so that the text sorts in time order. The numbers, the width, and the checks of the values use every row of the results, so filter_requested and filter_excluded never renumber an interval or hide a bad value; an interval with no rows in the results has no number. The grouping columns also remain in the output.

filter_requested

Only return rows with parameters that were specifically requested?

filter_excluded

Should excluded values be removed?

grpid_cols

For out_format = "cdisc", the grouping columns that prefix the interval number in PPGRPID: a named character vector of grouping columns of the PKNCAconc() object, in prefix order, other than the subject and the analyte. Each name is a column, and each value is the text written before that column's value, such as c(Part = "", Period = "P"). The text before a value and the values may not contain ".", and a value may not be empty, and distinct values of a column may not give the same text. NULL (the default) uses the grpid_cols of the PKNCAdata() object, and character() gives an interval-only identifier regardless of that default.

grpid_numeric

For out_format = "cdisc", the names of columns in grpid_cols whose values are whole numbers of at least 1 (such as the period), which are written as that number, so "01" becomes 1. A value that is not a finite whole number of at least 1 is an error. NULL (the default) uses the grpid_numeric of the PKNCAdata() object for the columns of grpid_cols that it names, and character() means no columns regardless of that default.

out.format

Deprecated in favor of out_format

Value

A data.frame (or usually a tibble) of results