Crate pmcore 

PMcore is a framework for developing and running population pharmacokinetic algorithms.

Algorithm Types

Non-Parametric Algorithms

Represent the population distribution as a discrete set of support points with associated weights.

• NPAG (Non-Parametric Adaptive Grid)
• NPOD (Non-Parametric Optimal Design)
• NPMAP (Maximum a posteriori reweighting)

Parametric Algorithms (planned)

Represent the population distribution with a parametric form (e.g. a normal distribution) and estimate the parameters of that distribution. This family is not yet implemented; the API is present but calling it will panic until a solver (SAEM) is available.

Public Interface

PMcore centers on the estimation interface in estimation. Models are defined in model, configured with estimation::EstimationProblem, and then executed with the selected algorithm.

Data format

PMcore is heavily linked to [pharmsol], which provides the data structures and routines for handling pharmacokinetic data. The data is stored in a pharmsol::Data structure, and can either be read from a CSV file, using pharmsol::data::parse_pmetrics::read_pmetrics, or created dynamically using the [pharmsol::data::builder::SubjectBuilder].

Modules

algorithms

Provides the various algorithms used within the framework

bestdose

Dose optimization and forecasting (BestDose).

estimation

Estimation problems and the non-parametric and parametric fitting families.

iov

SDE-based Inter-Occasion Variability optimization. SDE-based Inter-Occasion Variability (IOV) analysis.

logs

Logging utilities.

model

Model-domain types: parameters, parameter spaces, and metadata.

prelude

A collection of commonly used items to simplify imports.

results

Result and summary types shared across algorithms.

Structs

HashMap

A hash map implemented with quadratic probing and SIMD lookup.

Type Aliases

Result

Result<T, Error>