Diagonal matrix generation diagonal matrix fit using separate univariate normals

[Generated automatically as a Tutorial summary]

Model Description

Name:: gen_indep_fit_indep
Title:: Diagonal matrix generation diagonal matrix fit using separate univariate normals
Author:: PoPy for PK/PD
Abstract:

One compartment model with absorption compartment and CL/V parametrisation.
This script uses a diagonal covariance matrix to generate the data and a diagonal covariance matrix to fit.
Note here the ‘diagonal matrix’ is implemented as two separate univariate normal distributions, which is equivalent.

Keywords:: dep_one_cmp_cl; one compartment model; diagonal matrix
Input Script:: gen_indep_fit_indep_tut.pyml
Diagram:

Comparison

True objective value

-2214.8455

Final fitted objective value

-2215.9681

Compare Main f[X]

No Main f[X] values to compare.

Compare Noise f[X]

No Noise f[X] values to compare.

Compare Variance f[X]

Name	Initial	Fitted	True	Abs. Error	Prop. Error
f[CL_isv]	0.01	0.197	0.2	2.82e-03	1.41%
f[V_isv]	0.01	0.117	0.1	1.74e-02	17.38%

Outputs

Fitted f[X] values (after fitting)

f[KA] = 0.3000
f[CL] = 3.0000
f[V] = 20.0000
f[PNOISE_STD] = 0.1000
f[ANOISE_STD] = 0.0500
f[CL_isv] = 0.1972
f[V_isv] = 0.1174

Generated data .csv file

Synthetic Data:: synthetic_data.csv

Gen and Fit Summaries

Inputs

True f[X] values (for simulation)

f[KA] = 0.3000
f[CL] = 3.0000
f[V] = 20.0000
f[PNOISE_STD] = 0.1000
f[ANOISE_STD] = 0.0500
f[CL_isv] = 0.2000
f[V_isv] = 0.1000

Starting f[X] values (before fitting)

f[KA] = 0.3000
f[CL] = 3.0000
f[V] = 20.0000
f[PNOISE_STD] = 0.1000
f[ANOISE_STD] = 0.0500
f[CL_isv] = 0.0100
f[V_isv] = 0.0100