Log Linear

class synergy.single.LogLinear(aggregation_function=<function median>, nan_inverses=False, **kwargs)

Bases: DoseResponseModel1D

A model that fits dose response curves as piecewise linear interpolations of E vs log(dose).

This model is useful for drugs whose dose response do not follow some known parametric equation.

E(d)

Return the model’s effect(s) at dose(s) d.

Parameters:

d (ArrayLike) – Doses

Return ArrayLike:

Effects at doses d

E_inv(E)

Return the dose(s) required to achieve effect(s) E.

Parameters:

E (ArrayLike) – Effects

Return ArrayLike:

Doses required to achieve effects E

static create_fit(d, E, aggregation_function=<function median>)

Factory method to build a log-linear model directly from data.

fit(d, E, **kwargs)

Fit the model to data.

Parameters:

• d (ArrayLike) – Doses

• E (ArrayLike) – Measured dose-response effect at doses d

• kwargs – Additional arguments to pass to the fitting function

property is_fit

True if the model has been fit to data.

property is_specified

True if all parameters are set.