1 /* ----------------------------------------------------------------------
2 * Copyright (C) 2010-2014 ARM Limited. All rights reserved.
4 * $Date: 19. March 2015
7 * Project: CMSIS DSP Library
10 * Description: Standard deviation of an array of Q15 type.
12 * Target Processor: Cortex-M4/Cortex-M3/Cortex-M0
14 * Redistribution and use in source and binary forms, with or without
15 * modification, are permitted provided that the following conditions
17 * - Redistributions of source code must retain the above copyright
18 * notice, this list of conditions and the following disclaimer.
19 * - Redistributions in binary form must reproduce the above copyright
20 * notice, this list of conditions and the following disclaimer in
21 * the documentation and/or other materials provided with the
23 * - Neither the name of ARM LIMITED nor the names of its contributors
24 * may be used to endorse or promote products derived from this
25 * software without specific prior written permission.
27 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
28 * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
29 * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
30 * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
31 * COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
32 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
33 * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
34 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
35 * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
36 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
37 * ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
38 * POSSIBILITY OF SUCH DAMAGE.
39 * -------------------------------------------------------------------- */
53 * @brief Standard deviation of the elements of a Q15 vector.
54 * @param[in] *pSrc points to the input vector
55 * @param[in] blockSize length of the input vector
56 * @param[out] *pResult standard deviation value returned here
60 * <b>Scaling and Overflow Behavior:</b>
63 * The function is implemented using a 64-bit internal accumulator.
64 * The input is represented in 1.15 format.
65 * Intermediate multiplication yields a 2.30 format, and this
66 * result is added without saturation to a 64-bit accumulator in 34.30 format.
67 * With 33 guard bits in the accumulator, there is no risk of overflow, and the
68 * full precision of the intermediate multiplication is preserved.
69 * Finally, the 34.30 result is truncated to 34.15 format by discarding the lower
70 * 15 bits, and then saturated to yield a result in 1.15 format.
78 q31_t sum
= 0; /* Accumulator */
79 q31_t meanOfSquares
, squareOfMean
; /* square of mean and mean of square */
80 uint32_t blkCnt
; /* loop counter */
81 q63_t sumOfSquares
= 0; /* Accumulator */
83 #ifndef ARM_MATH_CM0_FAMILY
85 /* Run the below code for Cortex-M4 and Cortex-M3 */
87 q31_t in
; /* input value */
88 q15_t in1
; /* input value */
97 blkCnt
= blockSize
>> 2u;
99 /* First part of the processing with loop unrolling. Compute 4 outputs at a time.
100 ** a second loop below computes the remaining 1 to 3 samples. */
103 /* C = (A[0] * A[0] + A[1] * A[1] + ... + A[blockSize-1] * A[blockSize-1]) */
104 /* Compute Sum of squares of the input samples
105 * and then store the result in a temporary variable, sum. */
106 in
= *__SIMD32(pSrc
)++;
107 sum
+= ((in
<< 16) >> 16);
109 sumOfSquares
= __SMLALD(in
, in
, sumOfSquares
);
110 in
= *__SIMD32(pSrc
)++;
111 sum
+= ((in
<< 16) >> 16);
113 sumOfSquares
= __SMLALD(in
, in
, sumOfSquares
);
115 /* Decrement the loop counter */
119 /* If the blockSize is not a multiple of 4, compute any remaining output samples here.
120 ** No loop unrolling is used. */
121 blkCnt
= blockSize
% 0x4u
;
125 /* C = (A[0] * A[0] + A[1] * A[1] + ... + A[blockSize-1] * A[blockSize-1]) */
126 /* Compute Sum of squares of the input samples
127 * and then store the result in a temporary variable, sum. */
129 sumOfSquares
= __SMLALD(in1
, in1
, sumOfSquares
);
132 /* Decrement the loop counter */
136 /* Compute Mean of squares of the input samples
137 * and then store the result in a temporary variable, meanOfSquares. */
138 meanOfSquares
= (q31_t
)(sumOfSquares
/ (q63_t
)(blockSize
- 1));
140 /* Compute square of mean */
141 squareOfMean
= (q31_t
) ((q63_t
)sum
* sum
/ (q63_t
)(blockSize
* (blockSize
- 1)));
143 /* mean of the squares minus the square of the mean. */
144 /* Compute standard deviation and store the result to the destination */
145 arm_sqrt_q15(__SSAT((meanOfSquares
- squareOfMean
) >> 15, 16u), pResult
);
149 /* Run the below code for Cortex-M0 */
150 q15_t in
; /* input value */
158 /* Loop over blockSize number of values */
163 /* C = (A[0] * A[0] + A[1] * A[1] + ... + A[blockSize-1] * A[blockSize-1]) */
164 /* Compute Sum of squares of the input samples
165 * and then store the result in a temporary variable, sumOfSquares. */
167 sumOfSquares
+= (in
* in
);
169 /* C = (A[0] + A[1] + A[2] + ... + A[blockSize-1]) */
170 /* Compute sum of all input values and then store the result in a temporary variable, sum. */
173 /* Decrement the loop counter */
177 /* Compute Mean of squares of the input samples
178 * and then store the result in a temporary variable, meanOfSquares. */
179 meanOfSquares
= (q31_t
)(sumOfSquares
/ (q63_t
)(blockSize
- 1));
181 /* Compute square of mean */
182 squareOfMean
= (q31_t
) ((q63_t
)sum
* sum
/ (q63_t
)(blockSize
* (blockSize
- 1)));
184 /* mean of the squares minus the square of the mean. */
185 /* Compute standard deviation and store the result to the destination */
186 arm_sqrt_q15(__SSAT((meanOfSquares
- squareOfMean
) >> 15, 16u), pResult
);
188 #endif /* #ifndef ARM_MATH_CM0_FAMILY */
194 * @} end of STD group