Pharmacometric modelling
Project description
Pharmpy is a library for pharmacometrics. It can be used as a regular python package, in R via reticulate or via its built in command line interface.
Pharmpy is architectured to be able to handle different types of model formats and data formats and exposes a model agnostic API.
Current features:
Parsing of many parts of a NONMEM model file
Parsing of NONMEM result files
CLI supporting dataset filtering, resampling, anonymization and viewing
Pharmpy is developed by the Uppsala University Pharmacometrics group.
0.11.0 (2020-11-20)
Method df.pharmpy.observations to extract observations from dataframe
Add ColumnTypes EVENT and DOSE
Add model.to_base_model to convert model to its raw base model form
New functions in modeling: remove_iiv, zero_order_elimination, comined_mm_fo_elimination and add_parameter
Split modeling.absorption_rate and error into multiple functions
Add calculations of AIC and BIC to ModelfitResults
Improved pretty printing
0.10.0 (2020-11-16)
modeling.create_rv_block
modeling.michaelis_menten_elimination
modeling.set_transit_compartments
First version of modelfit method
Add first version of bootstrap method
Add parameter estimates histograms to bootstrap report
Add automatic update of $SIZES PD when writing/updating NONMEM model
Additions to QAResults
NMTRanParseError replaced with ModelSyntaxError
Multiple bugfixes to frem and scm result calculations
0.9.0 (2020-10-26)
Add error_model function to the modeling module
Added more standard models for modeling.add_etas
Improve BootstrapResults
Add plots to bootstrap
Add support for the PHARMPYCONFIGPATH environment variable
Add QAResults and LinearizeResults classes
Bugfixes for some Windows specific issues
0.8.0 (2020-10-08)
Add basic modeling functions to the modeling module
modeling.add_etas
Improved bootstrap results generation and additional plots
Bugfix: Labelled OMEGAS could sometimes get wrong symbol names
0.7.0 (2020-09-28)
Add method reset_indices in Results to flatten multiindices. Useful from R.
absorption_rate can also set sequential zero first absorption
New functionsadd_lag_time and remove_lag_time in modeling
Add basic functions fix/unfix_parameter, update_source and read_model to modeling API
Updated reading of NONMEM results
Bugfixes in add_covariate_effects and absorption_rate
Fix crash in FREM results if no log option could be found in meta.yaml
0.6.0 (2020-09-18)
Add eta transformations: boxcox, t-dist and John Draper
Add results cdd and scm to CLI
Add different views for scm results
Add support for taking parameter names from comment in NONMEM model
Remove assumptions for symbols
Add modeling.absorption_rate to set 0th or first order absorption
Add update of $TABLE numbers
0.5.0 (2020-09-04)
Many bugfixes and improvements to NONMEM code record parser
Add calculation of symbolic and numeric eta and eps gradients, population and individulal prediction and wres for PRED models
Add option to use comments in NONMEM parameter records as names for parameters
Reading of ODE systems from NONMEM non-$DES models
Calculation of compartmental matrix and ODE system
New module ‘modeling’
Function in modeling and CLI to change ADVAN implicit compartmental models to explicit $DES
Function in modeling and CLI to add covariate effects
Functions for reading cdd and scm results from PsN runs
Many API updates
Extended CLI documentation
0.4.0 (2020-07-24)
Add categorical covariates to covariate effects plot in FREM
Better support for reading NONMEM code statements (PK and PRED)
Support for updating NONMEM code statements (PK and PRED)
Bugfixes for CLI
0.3.0 (2020-06-16)
New CLI command ‘data append’
Parameter names is now the index in Parameters.summary()
FREM postprocessing
Standardized results.yaml and results.csv
0.2.0 (2020-03-27)
First release
0.1.0 (2018-07-22)
Initial library development/testing directory structure.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for pharmpy_core-0.11.0-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | ef1a5a2172d7f44f79d4ff016b38cf8f6ae5f26b619c9c532363b4211a346898 |
|
MD5 | 3142a61e7f775af8f42cb14b38073d1f |
|
BLAKE2b-256 | 7704d21e9128cff8bc8007c125b4f180f5101502f85c51cfd8187dd296ea3903 |