blob: ff4e7df55820e03305ff1624c44ff22f9fffb36e (
plain) (
blame)
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
|
// Copyright Vespa.ai. Licensed under the terms of the Apache 2.0 license. See LICENSE in the project root.
#pragma once
#include <vespa/searchlib/common/feature.h>
#include <cmath>
namespace search::features {
/**
* This class is used to calculate a logarithmic-shaped function that goes from 1 to 0.
* The function is:
* logscale(x, m, s) = (x > m ? 0 : (( log(m + s) - log(x + s)) / (log(m + s) - log(s)))),
* where m specifies for which x the function should output 0 (max parameter),
* and s controls the shape of the function (scale parameter).
*
* If you decide a value for x for when the function should output 0.5,
* s can be calculated as -x^2/(2x - m).
**/
class LogarithmCalculator {
private:
feature_t _m;
feature_t _s;
feature_t _maxLog;
feature_t _minLog;
feature_t _divMult;
public:
/**
* Creates a calculator for the given values for m (max) and s (scale).
**/
LogarithmCalculator(feature_t m, feature_t s) :
_m(m),
_s(s),
_maxLog(std::log(_m + _s)),
_minLog(std::log(_s)),
_divMult(1.0 / (_maxLog - _minLog))
{
}
/**
* Calculate the function for the given x.
**/
feature_t get(feature_t x) const {
if (x > _m) x = _m;
if (x < 0) x = 0;
return (_maxLog - std::log(x + _s)) * _divMult;
}
/**
* Calculate the scale parameter to use if the function should output 0.5
* for the given x and max parameter.
*/
static feature_t getScale(feature_t x, feature_t m) {
return (x * x) / (m - 2*x);
}
};
}
|
