Parameter¶

lumicks.pylake.fitting.parameters.Parameter

class Parameter(value=0.0, lower_bound=- inf, upper_bound=inf, fixed=False, shared=False, unit=None, stderr=None)¶

A single model parameter, usually part of a Params dictionary.

Examples

import lumicks.pylake as lk
fit = lk.FdFit(lk.ewlc_odijk_distance("my_model"))

print(fit.params)  # Prints the model parameters
parameter = fit["my_model/Lc"] = 5  # parameter is of the type `Parameter` here.

# You can extract and/or modify fitting bounds
lower_bound = parameter.lower_bound

# Or read out the fitting error after the model has been fitted.
print(f"fitting error Lc: {parameter.stderr}")

ci(percentile=0.95, dof=1)¶

Calculate confidence intervals

Parameters:

percentile (float) – 1 - Significance level
dof (float) – Degrees of freedom (should be 1 for single parameter CIs).

fixed¶: Parameter is held fixed during fitting.

lower_bound¶: Lower bound used when fitting.

profile¶

Profile likelihood result.

Profile likelihood estimates confidence intervals for the model parameters. These confidence intervals can be used to assess whether a parameter can reliably be estimated from the data. See also: profile_likelihood().

shared¶

Parameter is shared between all sub-models.

Some parameters are not expected to be different between sub-models.

stderr¶

Standard error of this parameter.

Standard errors are calculated after fitting the model. These asymptotic errors are based on the fitted parameter values and model sensitivities.

Note

These errors may be inaccurate in the presence of model non-identifiability. See also: profile_likelihood().

unit¶: Unit of this parameter.

upper_bound¶: Upper bound used when fitting.

value¶: Parameter value.