Methods for imputation of data with PKNCA
Arguments
- conc
Measured concentrations
- time
Time of the measurement of the concentrations
- start
The start time of the interval
- ...
ignored
- options
List of changes to the default PKNCA options (see
PKNCA.options()
)- end
The end time of the interval
- conc.group
All concentrations measured for the group
- time.group
Time of all concentrations measured for the group
- max_shift
The maximum amount of time to shift a concentration forward (defaults to 5% of the interval duration, i.e.
0.05*(end - start)
, ifis.finite(end)
, and whenis.infinite(end)
, defaults to 5% of the time from start tomax(time)
)
Value
A data.frame with one column named conc with imputed concentrations and one column named time with the times.
Functions
PKNCA_impute_method_start_conc0()
: Add a new concentration of 0 at the start time, even if a nonzero concentration exists at that time (usually used with single-dose data)PKNCA_impute_method_start_cmin()
: Add a new concentration of the minimum during the interval at the start time (usually used with multiple-dose data)PKNCA_impute_method_start_predose()
: Shift a predose concentration to become the time zero concentration (only if a time zero concentration does not exist)