blob: 26c44ebc8c8c10ef957420763b7d7c4a57b5f8d3 (
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
|
#pragma once
#include <array>
#include <string>
#include <map>
/**
* About:
* This is implementation of Independent two-sample t-test
* Read about it on https://en.wikipedia.org/wiki/Student%27s_t-test (Equal or unequal sample sizes, equal variance)
*
* Usage:
* It's it used to assume with some level of confidence that two distributions don't differ.
* Values can be added with t_test.add(0/1, value) and after compared and reported with compareAndReport().
*/
class StudentTTest
{
private:
struct DistributionData
{
size_t size = 0;
double sum = 0;
double squares_sum = 0;
void add(double value)
{
++size;
sum += value;
squares_sum += value * value;
}
double avg() const
{
return sum / static_cast<double>(size);
}
double var() const
{
return (squares_sum - (sum * sum / size)) / static_cast<double>(size - 1);
}
void clear()
{
size = 0;
sum = 0;
squares_sum = 0;
}
};
std::array<DistributionData, 2> data {};
public:
void clear();
void add(size_t distribution, double value);
/// Confidence_level_index can be set in range [0, 5]. Corresponding values can be found above. TODO: Trash - no separation of concepts in code.
std::pair<bool, std::string> compareAndReport(size_t confidence_level_index = 5) const;
};
|
