minuit2 0.4.1

Pure Rust port of CERN Minuit2 parameter optimization engine
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
149
150
151

//! Public Migrad (variable-metric) minimizer API.
//!
//! `MnMigrad` is the user-facing entry point for quasi-Newton minimization
//! using the Davidon-Fletcher-Powell (DFP) update of the inverse Hessian.
//! Uses a builder pattern to configure parameters, then call `minimize()`.

pub mod builder;
pub mod minimizer;
pub mod seed;

use crate::application::{DEFAULT_TOLERANCE, default_max_fcn};
use crate::fcn::{FCN, FCNGradient};
use crate::minimum::FunctionMinimum;
use crate::mn_fcn::MnFcn;
use crate::strategy::MnStrategy;
use crate::user_parameters::MnUserParameters;

/// Builder for configuring and running Migrad minimization.
pub struct MnMigrad {
    params: MnUserParameters,
    strategy: MnStrategy,
    max_fcn: Option<usize>,
    tolerance: f64,
}

impl MnMigrad {
    /// Create a new Migrad minimizer with default strategy.
    /// Create a new Migrad minimizer with default settings.
    pub fn new() -> Self {
        Self {
            params: MnUserParameters::new(),
            strategy: MnStrategy::default(),
            max_fcn: None,
            tolerance: DEFAULT_TOLERANCE,
        }
    }

    /// Set strategy level (0=low, 1=medium, 2=high).
    /// Set the optimization strategy level.
    pub fn with_strategy(mut self, level: u32) -> Self {
        self.strategy = MnStrategy::new(level);
        self
    }

    /// Add a free parameter.
    pub fn add(mut self, name: impl Into<String>, value: f64, error: f64) -> Self {
        self.params.add(name, value, error);
        self
    }

    /// Add a parameter with both bounds.
    pub fn add_limited(
        mut self,
        name: impl Into<String>,
        value: f64,
        error: f64,
        lower: f64,
        upper: f64,
    ) -> Self {
        self.params.add_limited(name, value, error, lower, upper);
        self
    }

    /// Add a parameter with lower bound only.
    pub fn add_lower_limited(
        mut self,
        name: impl Into<String>,
        value: f64,
        error: f64,
        lower: f64,
    ) -> Self {
        self.params.add_lower_limited(name, value, error, lower);
        self
    }

    /// Add a parameter with upper bound only.
    pub fn add_upper_limited(
        mut self,
        name: impl Into<String>,
        value: f64,
        error: f64,
        upper: f64,
    ) -> Self {
        self.params.add_upper_limited(name, value, error, upper);
        self
    }

    /// Add a constant parameter.
    pub fn add_const(mut self, name: impl Into<String>, value: f64) -> Self {
        self.params.add_const(name, value);
        self
    }

    /// Fix parameter by index.
    pub fn fix(mut self, ext: usize) -> Self {
        self.params.fix(ext);
        self
    }

    /// Set maximum number of function calls. Default = 200 + 100*n + 5*n^2.
    pub fn max_fcn(mut self, max: usize) -> Self {
        self.max_fcn = Some(max);
        self
    }

    /// Set tolerance (relative to error_def). Default = 0.1.
    pub fn tolerance(mut self, tol: f64) -> Self {
        self.tolerance = tol;
        self
    }

    /// Run the minimization with numerical gradients (default).
    pub fn minimize(&self, fcn: &dyn FCN) -> FunctionMinimum {
        let n = self.params.variable_parameters();
        let max_fcn = self.max_fcn.unwrap_or_else(|| default_max_fcn(n));
        let trafo = self.params.trafo().clone();

        let mn_fcn = MnFcn::new(fcn, &trafo);
        minimizer::VariableMetricMinimizer::minimize(
            &mn_fcn,
            &trafo,
            &self.strategy,
            max_fcn,
            self.tolerance,
        )
    }

    /// Run the minimization with user-provided analytical gradients.
    ///
    /// Uses the analytical gradients provided by `FCNGradient::gradient()`.
    /// This typically requires fewer function evaluations than numerical differentiation.
    pub fn minimize_grad(&self, fcn: &dyn FCNGradient) -> FunctionMinimum {
        let n = self.params.variable_parameters();
        let max_fcn = self.max_fcn.unwrap_or_else(|| default_max_fcn(n));
        let trafo = self.params.trafo().clone();

        minimizer::VariableMetricMinimizer::minimize_with_gradient(
            fcn,
            &trafo,
            &self.strategy,
            max_fcn,
            self.tolerance,
        )
    }
}

impl Default for MnMigrad {
    fn default() -> Self {
        Self::new()
    }
}