Single TAC Fitting¶

Use kinepet.fit_one_tac() when you have one curve.

fit = kinepet.fit_one_tac(
    tac,
    time,
    aif,
    model="irr",
    fit_vb=True,
)

tac, time, aif, and optional weights must all be one-dimensional arrays with the same length.

The options match fit_tacs(), except there is no threads argument because only one TAC is fitted.

Accessing results¶

fit_one_tac() returns SingleTacFitResult. Fields are scalars instead of arrays:

print(fit.K1)
print(fit.k2)
print(fit.k3)
print(fit.vB)
print(fit.Ki)
print(fit.rmse)
print(fit.status)
print(fit.status_code)

For model="rev", fit.k4 contains the reversible exchange parameter. For model="irr", fit.k4 is None.

Round trip with evaluate_model¶

tac = kinepet.evaluate_model(
    time,
    aif,
    model="rev",
    K1=0.12,
    k2=0.16,
    k3=0.09,
    k4=0.04,
    vB=0.05,
)

fit = kinepet.fit_one_tac(tac, time, aif, model="rev", max_iterations=200)