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
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
|
/* ----------------------------------------------------------------------
* Project: CMSIS DSP Library
* Title: arm_logsumexp_f16.c
* Description: LogSumExp
*
* $Date: 23 April 2021
* $Revision: V1.9.0
*
* Target Processor: Cortex-M and Cortex-A cores
* -------------------------------------------------------------------- */
/*
* Copyright (C) 2010-2021 ARM Limited or its affiliates. All rights reserved.
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed under the Apache License, Version 2.0 (the License); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an AS IS BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "dsp/statistics_functions_f16.h"
#if defined(ARM_FLOAT16_SUPPORTED)
#include <limits.h>
#include <math.h>
/**
* @addtogroup LogSumExp
* @{
*/
/**
* @brief Computation of the LogSumExp
*
* In probabilistic computations, the dynamic of the probability values can be very
* wide because they come from gaussian functions.
* To avoid underflow and overflow issues, the values are represented by their log.
* In this representation, multiplying the original exp values is easy : their logs are added.
* But adding the original exp values is requiring some special handling and it is the
* goal of the LogSumExp function.
*
* If the values are x1...xn, the function is computing:
*
* ln(exp(x1) + ... + exp(xn)) and the computation is done in such a way that
* rounding issues are minimised.
*
* The max xm of the values is extracted and the function is computing:
* xm + ln(exp(x1 - xm) + ... + exp(xn - xm))
*
* @param[in] *in Pointer to an array of input values.
* @param[in] blockSize Number of samples in the input array.
* @return LogSumExp
*
*/
#if defined(ARM_MATH_MVE_FLOAT16) && !defined(ARM_MATH_AUTOVECTORIZE)
#include "arm_helium_utils.h"
#include "arm_vec_math_f16.h"
float16_t arm_logsumexp_f16(const float16_t *in, uint32_t blockSize)
{
float16_t maxVal;
const float16_t *pIn;
int32_t blkCnt;
_Float16 accum=0.0f16;
_Float16 tmp;
arm_max_no_idx_f16((float16_t *) in, blockSize, &maxVal);
blkCnt = blockSize;
pIn = in;
f16x8_t vSum = vdupq_n_f16(0.0f16);
blkCnt = blockSize >> 3;
while(blkCnt > 0)
{
f16x8_t vecIn = vld1q(pIn);
f16x8_t vecExp;
vecExp = vexpq_f16(vsubq_n_f16(vecIn, maxVal));
vSum = vaddq_f16(vSum, vecExp);
/*
* Decrement the blockSize loop counter
* Advance vector source and destination pointers
*/
pIn += 8;
blkCnt --;
}
/* sum + log */
accum = vecAddAcrossF16Mve(vSum);
blkCnt = blockSize & 0x7;
while(blkCnt > 0)
{
tmp = *pIn++;
accum += (_Float16)expf((float32_t)((_Float16)tmp - (_Float16)maxVal));
blkCnt--;
}
accum = (_Float16)maxVal + (_Float16)logf((float32_t)accum);
return (accum);
}
#else
float16_t arm_logsumexp_f16(const float16_t *in, uint32_t blockSize)
{
_Float16 maxVal;
_Float16 tmp;
const float16_t *pIn;
uint32_t blkCnt;
_Float16 accum;
pIn = in;
blkCnt = blockSize;
maxVal = *pIn++;
blkCnt--;
while(blkCnt > 0)
{
tmp = *pIn++;
if (tmp > maxVal)
{
maxVal = tmp;
}
blkCnt--;
}
blkCnt = blockSize;
pIn = in;
accum = 0;
while(blkCnt > 0)
{
tmp = *pIn++;
accum += (_Float16)expf((float32_t)((_Float16)tmp - (_Float16)maxVal));
blkCnt--;
}
accum = (_Float16)maxVal + (_Float16)logf((float32_t)accum);
return(accum);
}
#endif /* defined(ARM_MATH_MVEF) && !defined(ARM_MATH_AUTOVECTORIZE) */
/**
* @} end of LogSumExp group
*/
#endif /* #if defined(ARM_FLOAT16_SUPPORTED) */
|
