Create a PopED database for an IQRmodel
IQRpopEDdb.Rd
Create a PopED database for an IQRmodel
IQRpopEDdb(model, design, modelSpec, transform_pop = FALSE, ...)
Arguments
- model
IQRmodel
object- design
list to define the design, can contain elements
time
( observation time points),dosing
(the dosing scheme),groupsize
(number of subjects per group). See details and example.- modelSpec
list to define model specification. See example. See
IQRnlmeEst
for a full description of the list structure.- transform_pop
Define the reference frame for the population parameters. If TRUE, the parameters are transformed according to the model specification and standard errors are reported on the transformed scale. If FALSE, standard error as reported on the original, untransformed scale.
- ...
Arguments for create.poped.database
Value
A PopED database object, see create.poped.database
.
Details
The IQRpopEDdb
function uses the model, design and model specification
input to internally generate model prediction functions, residual error
functions and parameterization functions as being required
by create.poped.database
. Subsequently, it calls
create.poped.database()
with the appropriate arguments. Additional
arguments can be passed using the ...
argument.
The design
argument defines the experimental design by a list of
observation time points,
dosing scheme,
group size.
The typical design definition looks like this:
design <- list(
time = list(
group1 = list(
OUTPUT1 = c(1, 2, 5, 10),
OUTPUT2 = c(1, 5, 20)
),
group2 = list(
OUTPUT1 = c(1, 2, 5, 10),
OUTPUT2 = c(1, 5, 20)
)
),
dosing = list(
group1 = IQRdosing(
TIME = 0, ADM = 1, AMT = 40, TINF = 0
),
group2 = . IQRdosing(
TIME = 0, ADM = 1, AMT = 40, TINF = 0
)),
groupsize = list(
group1 = 12, group2 = 3
)
)
Groups names do not have to follow the scheme "groupx", but can be arbitrary names.
The element time is defined per group and per model output. Vectors of time points are expected. They can have different length. If the observation time points are the same for all groups, it is sufficient to define them for exactly one group.
The element dosing is defined per group. See IQRdosing
for the correct specification of a dosing scheme.
The resulting dosing data frames for each group must have equal
dimensions. Sometimes this must be enforced by introducing zero amounts by
hand. If the dosing is the same for all groups, it is sufficient to define
dosing for exactly one group.
The element groupsize defines the number of subjects in each group.
The names found in the argument groupsize
define the available groups.
They are used if time
or dosing
have been defined for one
group only to duplicate the element for each available group.
NOTE ABOUT COVARIATES
By default, the AMT columns of the dosings are added as covariates such that popED can optimize the amounts.
Other covariates are not yet implemented and have to be specified manually via ...
See also
PopED functions: model_prediction
,
plot_model_prediction
,
evaluate_design
Other Optimal Experiment Design:
evaluate_IQRpopEDdb()
,
sim_IQRpopEDdb()