lineprior 0.9.0

Domain-agnostic action priors built from historical action sequences.
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

use crate::model::PriorBook;
use crate::score::entropy_bits;
use serde::Serialize;

/// Entropy of one state's prior distribution, in bits. Low entropy means
/// one action dominates and the prior is likely useful; high entropy
/// means many actions compete and a fallback search may be safer.
#[derive(Debug, Clone, Serialize)]
pub struct StateEntropy {
    pub state: String,
    pub entropy_bits: f64,
    pub num_actions: usize,
}

/// Aggregate statistics over an entire prior book, for the CLI `summary` command.
#[derive(Debug, Clone, Serialize)]
pub struct SummaryReport {
    pub num_states: usize,
    pub num_action_entries: usize,
    pub avg_confidence: f64,
    pub avg_entropy_bits: f64,
}

/// Per-state entropy, sorted lexicographically by state for determinism.
pub fn state_entropy(book: &PriorBook) -> Vec<StateEntropy> {
    let mut result: Vec<StateEntropy> = book
        .entries
        .iter()
        .map(|(state, actions)| {
            let priors: Vec<f64> = actions.iter().map(|a| a.prior).collect();
            StateEntropy {
                state: state.clone(),
                entropy_bits: entropy_bits(&priors),
                num_actions: actions.len(),
            }
        })
        .collect();
    result.sort_by(|a, b| a.state.cmp(&b.state));
    result
}

/// Summarizes a whole prior book: how many states/actions it covers and
/// how confident/decisive it tends to be.
pub fn summarize(book: &PriorBook) -> SummaryReport {
    let num_states = book.entries.len();
    let num_action_entries: usize = book.entries.values().map(Vec::len).sum();

    if num_action_entries == 0 {
        return SummaryReport {
            num_states,
            num_action_entries: 0,
            avg_confidence: 0.0,
            avg_entropy_bits: 0.0,
        };
    }

    let total_confidence: f64 = book
        .entries
        .values()
        .flat_map(|actions| actions.iter())
        .map(|a| a.confidence)
        .sum();
    let entropies = state_entropy(book);
    let total_entropy: f64 = entropies.iter().map(|e| e.entropy_bits).sum();

    SummaryReport {
        num_states,
        num_action_entries,
        avg_confidence: total_confidence / num_action_entries as f64,
        avg_entropy_bits: total_entropy / num_states.max(1) as f64,
    }
}

#[cfg(test)]
mod tests {
    use super::*;
    use crate::model::PriorAction;
    use std::collections::HashMap;

    fn action(name: &str, prior: f64, confidence: f64) -> PriorAction {
        PriorAction {
            action: name.to_string(),
            count: 10,
            weighted_count: 10.0,
            success_rate: None,
            mean_score: None,
            prior,
            confidence,
        }
    }

    #[test]
    fn low_entropy_when_one_action_dominates() {
        let mut entries = HashMap::new();
        entries.insert(
            "s".to_string(),
            vec![action("a", 0.99, 0.9), action("b", 0.01, 0.9)],
        );
        let book = PriorBook {
            entries,
            ..Default::default()
        };
        let entropy = state_entropy(&book);
        assert!(entropy[0].entropy_bits < 0.2);
    }

    #[test]
    fn high_entropy_when_actions_compete_evenly() {
        let mut entries = HashMap::new();
        entries.insert(
            "s".to_string(),
            vec![
                action("a", 0.25, 0.9),
                action("b", 0.25, 0.9),
                action("c", 0.25, 0.9),
                action("d", 0.25, 0.9),
            ],
        );
        let book = PriorBook {
            entries,
            ..Default::default()
        };
        let entropy = state_entropy(&book);
        assert!((entropy[0].entropy_bits - 2.0).abs() < 1e-9);
    }

    #[test]
    fn summarize_averages_across_the_whole_book() {
        let mut entries = HashMap::new();
        entries.insert("s1".to_string(), vec![action("a", 1.0, 0.5)]);
        entries.insert("s2".to_string(), vec![action("b", 1.0, 1.0)]);
        let book = PriorBook {
            entries,
            ..Default::default()
        };
        let summary = summarize(&book);
        assert_eq!(summary.num_states, 2);
        assert_eq!(summary.num_action_entries, 2);
        assert!((summary.avg_confidence - 0.75).abs() < 1e-9);
    }
}