Pharmpy is an open-source software package for pharmacometric modeling. It has functionality ranging from reading and manipulating model files and datasets to full tools where subsequent results are collected and presented.

Features include:

• A model abstraction which splits a model into core components which Pharmpy understands and can manipulate: parameters, random variables, statements (including ODE system), dataset, and execution steps

• An abstraction for modelfit results which splits a parsed results into core components: e.g. OFV, parameter estimates, relative standard errors (RSEs), residuals, predictions

• Functions for manipulation of models and datasets in the modeling-module: e.g. change structural model, add time-after-dose column, add covariate effects

• Tools for automated model development: building various aspects (structural, iiv, iov, ruv, covariates, …) of PK, PKPD, TMDD and drug-metabolite models automatically

• Tools to aid model development in the tools-module: execution of models within Python/R scripts, bootstrap, comparison of estimation methods

• Simplify scripting of workflows. Makes it possible to run scripts including calls to long running tools multiple times without having to rerun already finished tool runs.

• Support for multiple estimation tools: parse NONMEM models, execute NONMEM, nlmixr2, and rxODE2 models, run all Pharmpy tools with NONMEM and some with nlmixr2

For more comprehensive information and documentation, see: https://pharmpy.github.io

Pharmpy can be used as a regular Python package, in R via the pharmr package, or via its built in command line interface.

If you have any questions or want to discuss with other users and developers in the Pharmpy community, you can go to our discussion page: https://github.com/pharmpy/pharmpy/discussions

Getting started

>>> from pharmpy.modeling import read_model
>>> from pharmpy.tools import load_example_modelfit_results
>>> model = load_example_model("pheno")
>>> model.parameters
            value  lower upper    fix
POP_CL   0.004693   0.00     ∞  False
POP_VC   1.009160   0.00     ∞  False
COVAPGR  0.100000  -0.99     ∞  False
IIV_CL   0.030963   0.00     ∞  False
IIV_VC   0.031128   0.00     ∞  False
SIGMA    0.013086   0.00     ∞  False
>>> res = load_example_modelfit_results("pheno")
>>> res.parameter_estimates
POP_CL     0.004696
POP_VC     0.984258
COVAPGR    0.158920
IIV_CL     0.029351
IIV_VC     0.027906
SIGMA      0.013241
Name: estimates, dtype: float64
>>>

# Get help
pharmpy -h

# Remove first ID from dataset and save new model using new dataset
pharmpy data filter run1.mod 'ID!=1'

# Run tool for selecting IIV structure
pharmpy run iivsearch run1.mod

Contact

This is the team behind Pharmpy

Please ask a question in an issue or contact one of the maintainers if you have any questions.