Quick hit demos

These are all miscellaneous features that may be difficult to find in other documentation.

library(mrgsolve)
library(dplyr)

$ operator for model object gets the parameter value

If our model parameters are

mod <- mrgsolve:::house()

param(mod)

 Model parameters (N=14):
 name value . name  value
 CL   1     | SEX   0    
 D1   2     | SEXCL 0.7  
 F1   1     | SEXVC 0.85 
 IC50 10    | VC    20   
 KA   1.2   | WT    70   
 KIN  100   | WTCL  0.75 
 KOUT 2     | WTVC  1

we can pick a parameter value with

mod$CL
[1] 1
mod$WT
[1] 70

Or slice off multiple parameters

i_want <- c("CL", "WT")
mod[i_want]
$CL
[1] 1

$WT
[1] 70

Model names

For programming with a model object, we can extract the names

names(mod)
$param
 [1] "CL"    "VC"    "KA"    "F1"    "D1"    "WTCL"  "WTVC"  "SEXCL" "SEXVC"
[10] "KIN"   "KOUT"  "IC50"  "WT"    "SEX"  

$init
[1] "GUT"  "CENT" "RESP"

$capture
[1] "DV" "CP"

$omega
[1] "..."

$sigma
[1] "..."

$omega_labels
$omega_labels[[1]]
[1] "ECL"   "EVC"   "EKA"   "EKOUT"


$sigma_labels
$sigma_labels[[1]]
[1] "EXPO"

or get all of the model elements in a list (output not shown)

my_model <- as.list(mod)

Zero all random effect variances on the fly

If your model has random effects, you can easily and temporarily zero them out.

mod <- modlib("popex") %>% update(end=96) %>% Req(DV,CL,V)
Building popex ... done.
omat(mod)
$...
      [,1] [,2] [,3]
ECL:   0.3  0.0  0.0
EV:    0.0  0.1  0.0
EKA:   0.0  0.0  0.5

It is easy to simulate either with or without the random effects in the simulation: this change can be made on the fly.

Use zero_re to make all random effect variances zero

mod %>% zero_re %>% omat
$...
      [,1] [,2] [,3]
ECL:     0    0    0
EV:      0    0    0
EKA:     0    0    0

By default, both OMEGA and SIGMA are zeroed. Check the arguments for zero_re to see how to selectively zero OMEGA or SIGMA.

Compare the population output

mod %>% ev(amt=100) %>% mrgsim(nid=20) %>% plot

with

mod %>% ev(amt=100) %>% zero_re %>% mrgsim(nid=20) %>% plot

Plot formulae

We commonly plot simulated output with a special plot method. By default, you get all compartments and output variables in the plot.

mod <- mrgsolve:::house() %>% ev(amt=100)
mod %>% mrgsim %>% plot

The plot can be customized with a formula selecting variables to plot. Other arguments

to lattice::xyplot can be passed as well.

mod %>% mrgsim %>% plot(CP+RESP ~ time, lty=2, col="firebrick")
Warning: In subset.data.frame(as.data.frame(x), ...) :
 extra arguments 'lty', 'col' will be disregarded

Get a data frame of simulated data

By default mrgsolve returns an object of simulated data (and other stuff)

out <- mrgsim(mod)

class(out)
[1] "mrgsims"
attr(,"package")
[1] "mrgsolve"

But you can get a data frame with

out <- mrgsim_df(mod)

class(out)
[1] "data.frame"

or

out <- mrgsim(mod,output="df")

out <- mrgsim(mod) %>% as_tibble()