set_initial_estimates#

pharmpy.modeling.set_initial_estimates(model, inits, move_est_close_to_bounds=False, strict=True)[source]#

Update initial parameter estimate for a model

Updates initial estimates of population parameters for a model. If the new initial estimates are out of bounds or NaN this function will raise.

Parameters:

model (Model) – Pharmpy model to update initial estimates
inits (pd.Series or dict) – Initial parameter estimates to update
move_est_close_to_bounds (bool) – Move estimates that are close to bounds. If correlation >0.99 the correlation will be set to 0.9, if variance is <0.001 the variance will be set to 0.01.
strict (bool) – Whether all parameters in input need to exist in the model. Default is True

Returns:

Model – Pharmpy model object

Example

>>> from pharmpy.modeling import load_example_model, set_initial_estimates
>>> from pharmpy.tools import load_example_modelfit_results
>>> model = load_example_model("pheno")
>>> results = load_example_modelfit_results("pheno")
>>> model.parameters.inits  
{'POP_CL': 0.00469307, 'POP_VC': 1.00916, 'COVAPGR': 0.1, 'IIV_CL': 0.0309626, 'IIV_VC': ...}
>>> model = set_initial_estimates(model, results.parameter_estimates)
>>> model.parameters.inits  
{'POP_CL': 0.00469555, 'POP_VC': 0.984258, 'COVAPGR': 0.15892, 'IIV_CL': 0.0293508, ...}
>>> model = load_example_model("pheno")
>>> model = set_initial_estimates(model, {'POP_CL': 2.0})
>>> model.parameters['POP_CL']
Parameter("POP_CL", 2.0, lower=0.0, upper=∞, fix=False)