1 /* ----------------------------------------------------------------------
2 * Project: CMSIS DSP Library
3 * Title: arm_rms_f16.c
4 * Description: Root mean square value of the elements of a floating-point vector
5 *
6 * $Date: 23 April 2021
7 * $Revision: V1.9.0
8 *
9 * Target Processor: Cortex-M and Cortex-A cores
10 * -------------------------------------------------------------------- */
11 /*
12 * Copyright (C) 2010-2021 ARM Limited or its affiliates. All rights reserved.
13 *
14 * SPDX-License-Identifier: Apache-2.0
15 *
16 * Licensed under the Apache License, Version 2.0 (the License); you may
17 * not use this file except in compliance with the License.
18 * You may obtain a copy of the License at
19 *
20 * www.apache.org/licenses/LICENSE-2.0
21 *
22 * Unless required by applicable law or agreed to in writing, software
23 * distributed under the License is distributed on an AS IS BASIS, WITHOUT
24 * WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
25 * See the License for the specific language governing permissions and
26 * limitations under the License.
27 */
28
29 #include "dsp/statistics_functions_f16.h"
30
31 #if defined(ARM_FLOAT16_SUPPORTED)
32
33
34 /**
35 @ingroup groupStats
36 */
37
38 /**
39 @defgroup RMS Root mean square (RMS)
40
41 Calculates the Root Mean Square of the elements in the input vector.
42 The underlying algorithm is used:
43
44 <pre>
45 Result = sqrt(((pSrc[0] * pSrc[0] + pSrc[1] * pSrc[1] + ... + pSrc[blockSize-1] * pSrc[blockSize-1]) / blockSize));
46 </pre>
47
48 There are separate functions for floating point, Q31, and Q15 data types.
49 */
50
51 /**
52 @addtogroup RMS
53 @{
54 */
55
56 /**
57 @brief Root Mean Square of the elements of a floating-point vector.
58 @param[in] pSrc points to the input vector
59 @param[in] blockSize number of samples in input vector
60 @param[out] pResult root mean square value returned here
61 @return none
62 */
63
64 #if defined(ARM_MATH_MVE_FLOAT16) && !defined(ARM_MATH_AUTOVECTORIZE)
65
arm_rms_f16(const float16_t * pSrc,uint32_t blockSize,float16_t * pResult)66 void arm_rms_f16(
67 const float16_t * pSrc,
68 uint32_t blockSize,
69 float16_t * pResult)
70 {
71 float16_t pow = 0.0f;
72
73 arm_power_f16(pSrc, blockSize, &pow);
74
75 /* Compute Rms and store the result in the destination */
76 arm_sqrt_f16(pow / (float16_t) blockSize, pResult);
77 }
78 #else
79
arm_rms_f16(const float16_t * pSrc,uint32_t blockSize,float16_t * pResult)80 void arm_rms_f16(
81 const float16_t * pSrc,
82 uint32_t blockSize,
83 float16_t * pResult)
84 {
85 uint32_t blkCnt; /* Loop counter */
86 _Float16 sum = 0.0f16; /* Temporary result storage */
87 _Float16 in; /* Temporary variable to store input value */
88
89 #if defined (ARM_MATH_LOOPUNROLL) && !defined(ARM_MATH_AUTOVECTORIZE)
90
91 /* Loop unrolling: Compute 4 outputs at a time */
92 blkCnt = blockSize >> 2U;
93
94 while (blkCnt > 0U)
95 {
96 /* C = A[0] * A[0] + A[1] * A[1] + ... + A[blockSize-1] * A[blockSize-1] */
97
98 in = *pSrc++;
99 /* Compute sum of squares and store result in a temporary variable, sum. */
100 sum += in * in;
101
102 in = *pSrc++;
103 sum += in * in;
104
105 in = *pSrc++;
106 sum += in * in;
107
108 in = *pSrc++;
109 sum += in * in;
110
111 /* Decrement loop counter */
112 blkCnt--;
113 }
114
115 /* Loop unrolling: Compute remaining outputs */
116 blkCnt = blockSize % 0x4U;
117
118 #else
119
120 /* Initialize blkCnt with number of samples */
121 blkCnt = blockSize;
122
123 #endif /* #if defined (ARM_MATH_LOOPUNROLL) */
124
125 while (blkCnt > 0U)
126 {
127 /* C = A[0] * A[0] + A[1] * A[1] + ... + A[blockSize-1] * A[blockSize-1] */
128
129 in = *pSrc++;
130 /* Compute sum of squares and store result in a temporary variable. */
131 sum += ( in * in);
132
133 /* Decrement loop counter */
134 blkCnt--;
135 }
136
137 /* Compute Rms and store result in destination */
138 arm_sqrt_f16(sum / (float16_t) blockSize, pResult);
139 }
140 #endif /* defined(ARM_MATH_MVEF) && !defined(ARM_MATH_AUTOVECTORIZE) */
141
142 /**
143 @} end of RMS group
144 */
145
146 #endif /* #if defined(ARM_FLOAT16_SUPPORTED) */
147
148