sonic-uart/sliwin.cpp at main · ricynlee/sonic-uart · GitHub

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
#include "sliwin.hpp"
#include <immintrin.h>
#include <cmath>

///////////////////////////////////////////////////////////////////////
// sliwin
///////////////////////////////////////////////////////////////////////
sliwin::sliwin(size_t n) {
    if (n==0)
        throw "sliwin _size 0";
    _oldest = 0;
    _capacity = n;
    _size = 0;
#if defined(__AVX__)
    n = (n+7)/8*8; // ceiled to 8-float-aligned
    __m256* ptr = (__m256*)_mm_malloc(n*sizeof(float), 32);
    for (size_t i=0; i<n/8; i++) {
        ptr[i] = _mm256_setzero_ps();
    }
    _data = (float*)ptr;
#else
    _data = new float[n];
#endif
}

sliwin::~sliwin() {
#if defined(__AVX__)
    _mm_free(_data);
#else
    delete[] _data;
#endif
}

void sliwin::slide(const float e) {
    _data[_oldest] = e;
    _oldest = (_oldest+1) % _capacity;
    if (_size<_capacity)
        _size++;
}

float& sliwin::operator[](size_t i) { // [0] is latest
    if (i>=_size)
        throw "out of range";
    if (_size<_capacity)
        return _data[_size-1-i];
    else
        return _data[(_oldest+_capacity-1-i) % _capacity];
}

bool sliwin::filled(void) const {
    return (_size>=_capacity);
}

size_t sliwin::size(void) const {
    return _size;
}

///////////////////////////////////////////////////////////////////////
// sliwin_sum
///////////////////////////////////////////////////////////////////////
sliwin_sum::sliwin_sum(size_t n):sliwin(n) {
    _sum = 0.0f;
}

float sliwin_sum::slide(const float e) {
    _sum += e;
    if (_size>=_capacity)
        _sum -= _data[_oldest];
    sliwin::slide(e);
    return _sum;
}

float sliwin_sum::sum(void) const {
    return _sum;
}

///////////////////////////////////////////////////////////////////////
// sliwin_mean
///////////////////////////////////////////////////////////////////////
sliwin_mean::sliwin_mean(size_t n):sliwin_sum(n) {
    // nothing else
}

float sliwin_mean::slide(const float e) {
    sliwin_sum::slide(e);
    _mean = sum() / size();
    return _mean;
}

float sliwin_mean::mean(void) const {
    return _mean;
}

///////////////////////////////////////////////////////////////////////
// sliwin_stdd
///////////////////////////////////////////////////////////////////////
sliwin_stdd::sliwin_stdd(size_t n):sliwin_mean(n) {
    // nothing else
}

float sliwin_stdd::slide(const float e) {
    sliwin_mean::slide(e);
    _stdd = 0.0f;
#if defined(__AVX__)
    const __m256* ptr = (const __m256*)_data;
    __m256 tmp;
    for (size_t i=0; i<(size()+7)/8; i++) {
        tmp = _mm256_set1_ps(mean());
        tmp = _mm256_sub_ps(ptr[i], tmp);
        tmp = _mm256_mul_ps(tmp, tmp);
        for (size_t j=0; j<8 && i*8+j<size(); j++) {
            _stdd += ((const float*)(&tmp))[j];
        }
    }
#else
    for (size_t i=0; i<size(); i++) {
        _stdd += powf(_data[i]-mean(), 2);
    }
#endif
    if (size()>1) {
        _stdd = sqrtf(_stdd/(size()-1));
    } else {
        _stdd = 0;
    }
    return _stdd;
}

float sliwin_stdd::stdd(void) const {
    return _stdd;
}