blob: e28823344779e1867ae7903323e73c2091fd5e72 (
plain)
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
|
// Variance
//
// variance - smv
//
// Calculates the moving variance
//
// Inputs: 1
// Outputs: 1
// Constants: 0
#include "ts.h"
typedef struct {
uint_t k;
} ts_variance_params_t;
ts_parameter_t const ts_variance_p = { .n_inputs = 1, .n_outputs = 1, .n_constants = 1, .params_size = sizeof(ts_variance_params_t) };
uint_t ts_variance_offset( ts_constant_t const *c )
{
( void )c;
return ( c ? ((ts_variance_params_t const *)c)->k > 0 ? ((ts_variance_params_t const *)c)->k - 1 : 0 : 0 );
}
int ts_variance(
fp_t * const *y, uint_t y_n,
fp_t const * const *x, uint_t x_n,
uint_t n, ts_constant_t const *c
)
{
TS_PROLOGUE( ts_variance_p );
if ( !c ) return -1;
TS_CHECK_INPUT(x, x_n, n);
uint_t k = ((ts_variance_params_t const *)c)->k;
if (k == 0 || n < k) return -1;
uint_t xi = 0;
uint_t yi = k;
fp_t sum = 0; // Sum of terms
fp_t sum2 = 0; // Sum of squared terms
fp_t mean; // Mean
fp_t mean2; // Mean of squares
for( ; xi < k; ++xi )
{
sum += x[0][xi];
sum2 += _ts_square( x[0][xi] );
mean = sum / ( xi + 1 );
mean2 = sum2 / ( xi + 1 );
}
fp_t var = mean2 - _ts_square( mean );
y[0][k - 1] = var >= 0 ? var : 0.0;
fp_t rk = _ts_reciprocal( k );
for( ; xi < n; ++xi, ++yi )
{
mean = mean + ( x[0][xi] - x[0][xi - k] ) * rk;
mean2 = mean2 + ( _ts_square( x[0][xi] ) - _ts_square( x[0][xi - k] ) ) * rk;
var = mean2 - _ts_square( mean );
y[0][yi] = var >= 0 ? var : 0.0;
}
return 0;
}
|