1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135

use crate::prelude::{algorithms::Algorithm, ipm::burke, output::NPResult, settings::Settings};
use anyhow::{Error, Result};
use pharmsol::prelude::{
    data::{Data, ErrorModel, ErrorType},
    simulator::{psi, Equation},
};

use ndarray::{Array1, Array2};

use super::{initialization, output::CycleLog};

/// Posterior probability algorithm
/// Reweights the prior probabilities to the observed data and error model
pub struct POSTPROB<E: Equation> {
    equation: E,
    psi: Array2<f64>,
    theta: Array2<f64>,
    w: Array1<f64>,
    objf: f64,
    cycle: usize,
    converged: bool,
    gamma: f64,
    error_type: ErrorType,
    data: Data,
    c: (f64, f64, f64, f64),
    #[allow(dead_code)]
    settings: Settings,
    cyclelog: CycleLog,
}

impl<E: Equation> Algorithm<E> for POSTPROB<E> {
    fn new(settings: Settings, equation: E, data: Data) -> Result<Box<Self>, anyhow::Error> {
        Ok(Box::new(Self {
            equation,
            psi: Array2::default((0, 0)),
            theta: Array2::default((0, 0)),
            w: Array1::default(0),
            objf: f64::INFINITY,
            cycle: 0,
            converged: false,
            gamma: settings.error.value,
            error_type: match settings.error.class.as_str() {
                "additive" => ErrorType::Add,
                "proportional" => ErrorType::Prop,
                _ => panic!("Error type not supported"),
            },
            c: settings.error.poly,
            settings,
            data,

            cyclelog: CycleLog::new(),
        }))
    }
    fn into_npresult(&self) -> NPResult<E> {
        NPResult::new(
            self.equation.clone(),
            self.data.clone(),
            self.theta.clone(),
            self.psi.clone(),
            self.w.clone(),
            self.objf,
            self.cycle,
            self.converged,
            self.settings.clone(),
            self.cyclelog.clone(),
        )
    }
    fn get_settings(&self) -> &Settings {
        &self.settings
    }

    fn get_data(&self) -> &Data {
        &self.data
    }

    fn get_prior(&self) -> Array2<f64> {
        initialization::sample_space(&self.settings, &self.data, &self.equation).unwrap()
    }

    fn likelihood(&self) -> f64 {
        self.objf
    }

    fn inc_cycle(&mut self) -> usize {
        0
    }

    fn get_cycle(&self) -> usize {
        0
    }

    fn set_theta(&mut self, theta: Array2<f64>) {
        self.theta = theta;
    }

    fn get_theta(&self) -> &Array2<f64> {
        &self.theta
    }

    fn psi(&self) -> &Array2<f64> {
        &self.psi
    }

    fn convergence_evaluation(&mut self) {}

    fn converged(&self) -> bool {
        true
    }

    fn evaluation(&mut self) -> Result<(), (Error, NPResult<E>)> {
        self.psi = psi(
            &self.equation,
            &self.data,
            &self.theta,
            &ErrorModel::new(self.c, self.gamma, &self.error_type),
            false,
            false,
        );
        (self.w, self.objf) = burke(&self.psi).expect("Error in IPM");
        Ok(())
    }

    fn condensation(&mut self) -> Result<(), (Error, NPResult<E>)> {
        Ok(())
    }
    fn optimizations(&mut self) -> Result<(), (Error, NPResult<E>)> {
        Ok(())
    }

    fn logs(&self) {}

    fn expansion(&mut self) -> Result<(), (Error, NPResult<E>)> {
        Ok(())
    }
}