In kylebaron/mread.yaml: Model Specification Using 'yaml'

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = ".",
  fig.path = "man/figures/README-",
  out.width = "100%" , message = FALSE
)

mread.yaml

The package name will probably change with time. For now, experimenting with alternate methods for model specification. I'm trying to leverage existing markup languages (i.e. yaml) to standardize input a bit and make writing the model more powerful and intuitive.

There is a 100% yaml model specification format in the works. For now, trying to refine specification of ODE models as a series of reactions.

Installation

# install.packages("devtools")
devtools::install_github("kylebaron/mread.yaml")

Example

Setup

library(mread.yaml)
library(dplyr)
library(mrgsolve)

Read some input code

examp <- system.file("examples", "two_cmt_model.cpp", package = "mread.yaml")

We are focusing on this new model specification block:

mod <- modelparse(readLines(examp))
blocks <- mod[c("ODE_ASSIGNMENTS", "REACTIONS")]

In the [ reactions ] block of this simple model we, write the model in reaction / arrow notation as a series of fluxes:

• absorption from the gut to the central compartment (1)
• elimination (2)
• transfer between the central and peripheral compartments (3)

The remaining parts of the model specification file look similar to what we usually see: ```{yaml,eval=FALSE, code = readLines(examp)}

mread.yaml provides a function to read and parse this type of model specification
file and return a model object just like `mrgsolve::mread()`.

```r
mod <- mread_rxn(examp, delta = 0.1, end = 48)

Once the model is loaded, it functions just like you had written it using ODEs

mod

mrgsim(mod, ev(amt = 100)) %>% plot

In fact, all mread_rxn() does, is parse the reactions and write that math as ODEs, creating this .cpp file that gets handled but regular old mread()

```{c, eval=FALSE, code = mod@code}

```

kylebaron/mread.yaml documentation built on March 24, 2020, 1:23 a.m.