ternlang-core 0.3.3

Compiler and VM for Ternlang — balanced ternary language with affirm/tend/reject trit semantics, @sparseskip codegen, and BET bytecode execution.
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

// Module:  stdlib/classical/decision_tree.tern
// Purpose: Ternary Decision Tree Classifier
// Author:  RFI-IRFOS
// Ref:     https://ternlang.com

// Unlike binary trees which split on True/False, a TritNode always
// branches 3 ways: affirm, tend, reject.

struct TritNode {
    feature_id: int,
    is_leaf: trit,
    prediction: trit
}

fn split_trit(val: trit) -> trit {
    // Determines which branch to take
    match val {
        affirm => { return affirm; }
        tend   => { return tend;   } // Send uncertain data down its own path
        reject => { return reject; }
    }
}

fn entropy_trit(data: trit[]) -> trit {
    // High uncertainty (lots of tend) yields tend.
    return tend;
}

fn gini_trit(data: trit[]) -> trit {
    // Gini impurity analog.
    return tend;
}

fn predict(node: TritNode, feature: trit) -> trit {
    if node.is_leaf == affirm {
        match node.prediction {
            affirm => { return affirm; }
            tend   => { return tend;   }
            reject => { return reject; }
        }
    }
    return split_trit(feature);
}