3  Read and write models

3.1 Read in a model file

Use the mread() function to read in a model coded in a file on the disk. For example, to read in a model for azithromycin population PK

mod <- mread("model/azithro.mod")

You can use any extension, but I’m preferring something like .mod or .txt to prevent your text editor from trying to reformat the model code.

You can update the model object on the fly by passing arguments to mread(). For example, to run the simulation out to one week and only have CP appear in the simulated output

mod <- mread("model/azithro.mod", end = 168, outvars = "CP")

To do dynamic capture on loading the model, pass the capture argument. For example, I want to get the clearance (CL) and the ETA associated with clearance (ETACL) into the simulation output

mod <- mread("model/azithro.mod", capture = "CL, ETACL")

To cache the compiled model, use mread_cache()

mod <- mread_cache("model/azithro.mod")

When you try to read the model again, it will load from the cache

mod <- mread_cache("model/azithro.mod")

. Loading model from cache.

To read in a model from the model library, run modlib(). For example, to read in a two-compartment model from the library

mod <- modlib("pk2")

To build the model in a specific location , use the soloc argument

mod <- modlib("irm1", soloc = "build")

. Loading model from cache.

This will build the model in the directory build off the current working directory.

You don’t have to compile the model

mod <- modlib("popex", compile = FALSE)

3.2 Write a model to file

With mrgsolve 1.5.1, you can also write a model object back to a file. The format of the file can be either native mrgsolve format or a yaml-formatted file. When the model object is written back to file, all of the updates that were made to the model object will be written to the file.

Once we have a model object

mod <- modlib("popex")

we can update it

mod <- update(mod, end = 72, outvars = "IPRED, CL, DV")
mod <- param(mod, TVCL = 1.2)

and then write the object out to yaml format with mwrite_yaml()

mwrite_yaml(mod, file = "popex.yaml")

Now, this model object is in a file called popex.yaml.

To read the model back in, use mread_yaml()

mod2 <- mread_yaml("popex.yaml")

You’ll see that the updated settings are retained in mod2

param(mod2)

. 
.  Model parameters (N=4):
.  name value . name value
.  TVCL 1.2   | TVV  24   
.  TVKA 0.5   | WT   70

mod2@end

. [1] 72

outvars(mod2)

. $cmt
. character(0)
. 
. $capture
. [1] "CL"    "IPRED" "DV"

In addition to writing the model object to yaml format, you can also write to native mrgsolve format with the mwrite_cpp() function

mwrite_cpp(mod2, file = "popex-updated.mod")