• Language: en

Depot + One compartment PK with BLQ

[Generated automatically as a Fitting summary]

Inputs

Description

Name:blq_pk
Title:Depot + One compartment PK with BLQ
Author:J.R. Hartley
Abstract:
Depot One Comp PK model, with BLQ (below level of quantification) observations.
Keywords:tutorial; pk; advan4; dep_two_cmp; blq
Input Script:blq_pk_fit.pyml
Input Data:synthetic_data.csv
Diagram:

Initial fixed effect estimates

f[KA] = 1.0000
f[CL] = 1.0000
f[V1] = 20.0000
f[KA_isv,CL_isv,V1_isv] = [
    [ 0.0500, 0.0100, 0.0100 ],
    [ 0.0100, 0.0500, 0.0100 ],
    [ 0.0100, 0.0100, 0.0500 ],
]
f[PNOISE] = 0.1000
f[ANOISE] = 0.0100

Outputs

Final objective value

-732.3632

which required N. iterations and took 268.74 seconds

Final fitted fixed effects

f[KA] = 0.2185
f[CL] = 1.9944
f[V1] = 50.4071
f[KA_isv,CL_isv,V1_isv] = [
    [ 0.0203, 0.0106, 0.0251 ],
    [ 0.0106, 0.0126, 0.0145 ],
    [ 0.0251, 0.0145, 0.1816 ],
]
f[PNOISE] = 0.1509
f[ANOISE] = 0.0100

Fitted parameter .csv files

Fixed Effects:fx_params.csv (fit)
Random Effects:rx_params.csv (fit)
Model params:mx_params.csv (fit)
State values:sx_params.csv (fit)
Predictions:px_params.csv (fit)

Plots

Dense sim plots

Alternatively see All dense_sim graph plots

Comparison

Compare Main f[X]

Variable Name Fitted Value Starting Value Prop Change Abs Change
f[KA] 0.2185 1.0000 0.7815 0.7815
f[CL] 1.9944 1.0000 0.9944 0.9944
f[V1] 50.4071 20.0000 1.5204 30.4071

Compare Noise f[X]

Variable Name Fitted Value Starting Value Prop Change Abs Change
f[PNOISE] 0.1509 0.1000 0.5089 0.0509

Compare Variance f[X]

Variable Name Fitted Value Starting Value Prop Change Abs Change
f[KA_isv] 0.0203 0.0500 0.5940 0.0297
f[KA_isv;CL_isv] 0.0106 0.0100 0.0639 0.0006
f[KA_isv;V1_isv] 0.0251 0.0100 1.5135 0.0151
f[CL_isv;KA_isv] 0.0106 0.0100 0.0639 0.0006
f[CL_isv] 0.0126 0.0500 0.7486 0.0374
f[CL_isv;V1_isv] 0.0145 0.0100 0.4461 0.0045
f[V1_isv;KA_isv] 0.0251 0.0100 1.5135 0.0151
f[V1_isv;CL_isv] 0.0145 0.0100 0.4461 0.0045
f[V1_isv] 0.1816 0.0500 2.6318 0.1316
Back to Top