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
152
153
154
155
156
157
158
159
160
161
162
163
164

//! A solver based on [Optuna](https://github.com/optuna/optuna).
use kurobako_core::epi::solver::{
    EmbeddedScriptSolver, EmbeddedScriptSolverFactory, EmbeddedScriptSolverRecipe,
};
use kurobako_core::problem::ProblemSpec;
use kurobako_core::registry::FactoryRegistry;
use kurobako_core::rng::ArcRng;
use kurobako_core::solver::{Solver, SolverFactory, SolverRecipe, SolverSpec};
use kurobako_core::trial::{EvaluatedTrial, IdGen, NextTrial};
use kurobako_core::Result;
use serde::{Deserialize, Serialize};
use structopt::StructOpt;

fn add_arg(args: &mut Vec<String>, key: &str, val: &str) {
    args.push(key.to_owned());
    args.push(val.to_owned());
}

#[allow(clippy::trivially_copy_pass_by_ref)]
fn is_false(b: &bool) -> bool {
    !(*b)
}

mod defaults {
    macro_rules! define {
        ($val_fn:ident, $pred_fn:ident, $type:ty, $val:expr) => {
            #[allow(clippy::ptr_arg, clippy::float_cmp)]
            pub fn $pred_fn(x: &$type) -> bool {
                x == &$val
            }

            pub fn $val_fn() -> $type {
                $val
            }
        };
    }

    define!(loglevel, is_loglevel, String, "warning".to_owned());
}

/// Recipe of `OptunaSolver`.
#[derive(Debug, Clone, StructOpt, Serialize, Deserialize)]
#[allow(missing_docs)]
#[structopt(rename_all = "kebab-case")]
pub struct OptunaSolverRecipe {
    /// Log level.
    #[structopt(
        long,
        default_value = "warning",
        possible_values = &["debug", "info", "warning", "error"]
    )]
    #[serde(skip_serializing_if = "defaults::is_loglevel")]
    #[serde(default = "defaults::loglevel")]
    pub loglevel: String,

    /// Sampler class name (e.g., "TPESampler").
    #[structopt(long)]
    #[serde(skip_serializing_if = "Option::is_none")]
    #[serde(default)]
    pub sampler: Option<String>,

    /// Sampler arguments (e.g., "{\"seed\": 10}").
    #[structopt(long)]
    #[serde(skip_serializing_if = "Option::is_none")]
    #[serde(default)]
    pub sampler_kwargs: Option<String>,

    /// Pruner class name (e.g., "MedianPruner").
    #[structopt(long)]
    #[serde(skip_serializing_if = "Option::is_none")]
    #[serde(default)]
    pub pruner: Option<String>,

    /// Pruner arguments (e.g., "{\"n_warmup_steps\": 10}").
    #[structopt(long)]
    #[serde(skip_serializing_if = "Option::is_none")]
    #[serde(default)]
    pub pruner_kwargs: Option<String>,

    /// Sets optimization direction to "maximize".
    ///
    /// The sign of all evaluated values ​​is reversed before being passed to Optuna.
    #[structopt(long)]
    #[serde(default, skip_serializing_if = "is_false")]
    pub maximize: bool,

    /// If this is `true`, `Trial.suggest_discrete_uniform()` is used for sampling discrete parameters instead of `Trial.suggest_int()`.
    #[structopt(long)]
    #[serde(default, skip_serializing_if = "is_false")]
    pub use_discrete_uniform: bool,
}
impl OptunaSolverRecipe {
    fn build_args(&self) -> Vec<String> {
        let mut args = Vec::new();
        add_arg(&mut args, "--loglevel", &self.loglevel);
        if let Some(v) = &self.sampler {
            add_arg(&mut args, "--sampler", v);
        }
        if let Some(v) = &self.sampler_kwargs {
            add_arg(&mut args, "--sampler-kwargs", v);
        }
        if let Some(v) = &self.pruner {
            add_arg(&mut args, "--pruner", v);
        }
        if let Some(v) = &self.pruner_kwargs {
            add_arg(&mut args, "--pruner-kwargs", v);
        }
        if self.maximize {
            args.push("--direction".to_owned());
            args.push("maximize".to_owned());
        }
        if self.use_discrete_uniform {
            args.push("--use-discrete-uniform".to_owned());
        }
        args
    }
}
impl SolverRecipe for OptunaSolverRecipe {
    type Factory = OptunaSolverFactory;

    fn create_factory(&self, registry: &FactoryRegistry) -> Result<Self::Factory> {
        let script = include_str!("../scripts/optuna_solver.py");
        let args = self.build_args();
        let recipe = EmbeddedScriptSolverRecipe {
            script: script.to_owned(),
            args,
        };
        let inner = track!(recipe.create_factory(registry))?;
        Ok(OptunaSolverFactory { inner })
    }
}

/// Factory of `OptunaSolver`.
#[derive(Debug)]
pub struct OptunaSolverFactory {
    inner: EmbeddedScriptSolverFactory,
}
impl SolverFactory for OptunaSolverFactory {
    type Solver = OptunaSolver;

    fn specification(&self) -> Result<SolverSpec> {
        track!(self.inner.specification())
    }

    fn create_solver(&self, rng: ArcRng, problem: &ProblemSpec) -> Result<Self::Solver> {
        let inner = track!(self.inner.create_solver(rng, problem))?;
        Ok(OptunaSolver { inner })
    }
}

/// Solver that uses [Optuna](https://github.com/optuna/optuna) as the backend.
#[derive(Debug)]
pub struct OptunaSolver {
    inner: EmbeddedScriptSolver,
}
impl Solver for OptunaSolver {
    fn ask(&mut self, idg: &mut IdGen) -> Result<NextTrial> {
        track!(self.inner.ask(idg))
    }

    fn tell(&mut self, trial: EvaluatedTrial) -> Result<()> {
        track!(self.inner.tell(trial))
    }
}