pharmsol 0.27.1

Rust library for solving analytic and ode-defined pharmacometric models.
Documentation 
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
136
137
138
139
140
141
142
143
144
145
146
147
148

//! Shows how to select different ODE solvers for the same declaration-first model.
//!
//! pharmsol wraps diffsol's solver families:
//!
//! - `Bdf`               — implicit multistep, best for stiff systems (default)
//! - `Sdirk(TrBdf2)`     — implicit single-step, good all-rounder
//! - `Sdirk(Esdirk34)`   — implicit single-step, higher accuracy
//! - `ExplicitRk(Tsit45)` — explicit Runge-Kutta, fastest for non-stiff systems
//!
//!     cargo run --release --example compare_solvers

use std::time::Instant;

use pharmsol::{prelude::*, Parameters};

// ── Model ──────────────────────────────────────────────────────────
// Two-compartment IV model. The solver is the only thing that changes
// between runs; the declaration-first `ode!` surface and the generated
// metadata stay the same.

fn two_cpt(solver: OdeSolver) -> equation::ODE {
    ode! {
        name: "two_cpt",
        params: [ke, kcp, kpc, v],
        states: [central, peripheral],
        outputs: [cp],
        routes: [
            bolus(load) -> central,
            infusion(iv) -> central,
        ],
        diffeq: |x, _p, _t, dx, _cov| {
            dx[central] = -ke * x[central] - kcp * x[central] + kpc * x[peripheral];
            dx[peripheral] = kcp * x[central] - kpc * x[peripheral];
        },
        out: |x, _p, _t, _cov, y| {
            y[cp] = x[central] / v;
        },
    }
    .with_solver(solver)
}

// ── Main ───────────────────────────────────────────────────────────

fn main() {
    // Run each solver and collect predictions
    let bdf = two_cpt(OdeSolver::Bdf);
    let tsit45 = two_cpt(OdeSolver::ExplicitRk(ExplicitRkTableau::Tsit45));
    let trbdf2 = two_cpt(OdeSolver::Sdirk(SdirkTableau::TrBdf2));
    let esdirk34 = two_cpt(OdeSolver::Sdirk(SdirkTableau::Esdirk34));

    // Both declarations resolve to the same shared input, so subject
    // authoring still uses one numeric index for the loading bolus and the
    // maintenance infusion.

    let subject = Subject::builder("id1")
        .bolus(0.0, 100.0, "load")
        .infusion(12.0, 200.0, "iv", 2.0)
        .missing_observation(0.5, "cp")
        .missing_observation(1.0, "cp")
        .missing_observation(2.0, "cp")
        .missing_observation(4.0, "cp")
        .missing_observation(8.0, "cp")
        .missing_observation(12.0, "cp")
        .missing_observation(12.5, "cp")
        .missing_observation(13.0, "cp")
        .missing_observation(14.0, "cp")
        .missing_observation(16.0, "cp")
        .missing_observation(24.0, "cp")
        .build();

    let parameters = Parameters::with_model(
        &bdf,
        [("ke", 0.1), ("kcp", 0.05), ("kpc", 0.03), ("v", 50.0)],
    )
    .expect("valid named parameters");

    let results: Vec<(&str, equation::ODE)> = vec![
        ("Bdf", bdf),
        ("Sdirk(TrBdf2)", trbdf2),
        ("Sdirk(Esdirk34)", esdirk34),
        ("ExplicitRk(Tsit45)", tsit45),
    ];

    // ── Run all solvers and collect results ───────────────────────
    let mut rows: Vec<(&str, u128, Vec<f64>)> = Vec::new();
    for (name, ode) in &results {
        let (preds, us) = timed(|| ode.estimate_predictions(&subject, &parameters).unwrap());
        let preds: Vec<f64> = preds.flat_predictions().to_vec();
        rows.push((name, us, preds));
    }

    let obs_times = [0.5, 1.0, 2.0, 4.0, 8.0, 12.0, 12.5, 13.0, 14.0, 16.0, 24.0];
    let ref_preds = &rows[0].2;

    // ── Summary ────────────────────────────────────────────────────
    println!();
    println!("  Solver Performance");
    println!("  {}", "─".repeat(48));
    for (i, (name, us, preds)) in rows.iter().enumerate() {
        let diff = if i == 0 {
            "(reference)".to_string()
        } else {
            let d = max_abs_diff(preds, ref_preds);
            format!("max \u{0394} {d:.2e}")
        };
        println!("  {:<22} {:>6} \u{00B5}s   {}", name, us, diff);
    }
    println!();

    // ── Predictions (long format) ──────────────────────────────────
    // Short labels for the column headers
    let labels: Vec<&str> = rows.iter().map(|(n, _, _)| *n).collect();
    let col_w = labels.iter().map(|l| l.len().max(8)).collect::<Vec<_>>();

    print!("  {:>6}", "t");
    for (label, width) in labels.iter().zip(&col_w) {
        print!("  {:>w$}", label, w = *width);
    }
    println!();

    print!("  {:>6}", "──────");
    for width in &col_w {
        print!("  {:─>w$}", "", w = *width);
    }
    println!();

    for (k, t) in obs_times.iter().enumerate() {
        print!("  {:>6.1}", t);
        for ((_, _, preds), width) in rows.iter().zip(&col_w) {
            print!("  {:>w$.4}", preds[k], w = *width);
        }
        println!();
    }
    println!();
}

fn timed<T>(f: impl FnOnce() -> T) -> (T, u128) {
    let t0 = Instant::now();
    let result = f();
    (result, t0.elapsed().as_micros())
}

fn max_abs_diff(a: &[f64], b: &[f64]) -> f64 {
    a.iter()
        .zip(b)
        .map(|(x, y)| (x - y).abs())
        .fold(0.0_f64, f64::max)
}