arm_var_f32.c 6.3 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204
  1. /* ----------------------------------------------------------------------
  2. * Copyright (C) 2010-2014 ARM Limited. All rights reserved.
  3. *
  4. * $Date: 12. March 2014
  5. * $Revision: V1.4.4
  6. *
  7. * Project: CMSIS DSP Library
  8. * Title: arm_var_f32.c
  9. *
  10. * Description: Variance of the elements of a floating-point vector.
  11. *
  12. * Target Processor: Cortex-M4/Cortex-M3/Cortex-M0
  13. *
  14. * Redistribution and use in source and binary forms, with or without
  15. * modification, are permitted provided that the following conditions
  16. * are met:
  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
  22. * distribution.
  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.
  26. *
  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. * ---------------------------------------------------------------------------- */
  40. #include "arm_math.h"
  41. /**
  42. * @ingroup groupStats
  43. */
  44. /**
  45. * @defgroup variance Variance
  46. *
  47. * Calculates the variance of the elements in the input vector.
  48. * The underlying algorithm is used:
  49. *
  50. * <pre>
  51. * Result = (sumOfSquares - sum<sup>2</sup> / blockSize) / (blockSize - 1)
  52. *
  53. * where, sumOfSquares = pSrc[0] * pSrc[0] + pSrc[1] * pSrc[1] + ... + pSrc[blockSize-1] * pSrc[blockSize-1]
  54. *
  55. * sum = pSrc[0] + pSrc[1] + pSrc[2] + ... + pSrc[blockSize-1]
  56. * </pre>
  57. *
  58. * There are separate functions for floating point, Q31, and Q15 data types.
  59. */
  60. /**
  61. * @addtogroup variance
  62. * @{
  63. */
  64. /**
  65. * @brief Variance of the elements of a floating-point vector.
  66. * @param[in] *pSrc points to the input vector
  67. * @param[in] blockSize length of the input vector
  68. * @param[out] *pResult variance value returned here
  69. * @return none.
  70. *
  71. */
  72. void arm_var_f32(
  73. float32_t * pSrc,
  74. uint32_t blockSize,
  75. float32_t * pResult)
  76. {
  77. float32_t sum = 0.0f; /* Temporary result storage */
  78. float32_t sumOfSquares = 0.0f; /* Sum of squares */
  79. float32_t in; /* input value */
  80. uint32_t blkCnt; /* loop counter */
  81. #ifndef ARM_MATH_CM0_FAMILY
  82. /* Run the below code for Cortex-M4 and Cortex-M3 */
  83. float32_t meanOfSquares, mean, squareOfMean; /* Temporary variables */
  84. if(blockSize == 1)
  85. {
  86. *pResult = 0;
  87. return;
  88. }
  89. /*loop Unrolling */
  90. blkCnt = blockSize >> 2u;
  91. /* First part of the processing with loop unrolling. Compute 4 outputs at a time.
  92. ** a second loop below computes the remaining 1 to 3 samples. */
  93. while(blkCnt > 0u)
  94. {
  95. /* C = (A[0] * A[0] + A[1] * A[1] + ... + A[blockSize-1] * A[blockSize-1]) */
  96. /* Compute Sum of squares of the input samples
  97. * and then store the result in a temporary variable, sum. */
  98. in = *pSrc++;
  99. sum += in;
  100. sumOfSquares += in * in;
  101. in = *pSrc++;
  102. sum += in;
  103. sumOfSquares += in * in;
  104. in = *pSrc++;
  105. sum += in;
  106. sumOfSquares += in * in;
  107. in = *pSrc++;
  108. sum += in;
  109. sumOfSquares += in * in;
  110. /* Decrement the loop counter */
  111. blkCnt--;
  112. }
  113. /* If the blockSize is not a multiple of 4, compute any remaining output samples here.
  114. ** No loop unrolling is used. */
  115. blkCnt = blockSize % 0x4u;
  116. while(blkCnt > 0u)
  117. {
  118. /* C = (A[0] * A[0] + A[1] * A[1] + ... + A[blockSize-1] * A[blockSize-1]) */
  119. /* Compute Sum of squares of the input samples
  120. * and then store the result in a temporary variable, sum. */
  121. in = *pSrc++;
  122. sum += in;
  123. sumOfSquares += in * in;
  124. /* Decrement the loop counter */
  125. blkCnt--;
  126. }
  127. /* Compute Mean of squares of the input samples
  128. * and then store the result in a temporary variable, meanOfSquares. */
  129. meanOfSquares = sumOfSquares / ((float32_t) blockSize - 1.0f);
  130. /* Compute mean of all input values */
  131. mean = sum / (float32_t) blockSize;
  132. /* Compute square of mean */
  133. squareOfMean = (mean * mean) * (((float32_t) blockSize) /
  134. ((float32_t) blockSize - 1.0f));
  135. /* Compute variance and then store the result to the destination */
  136. *pResult = meanOfSquares - squareOfMean;
  137. #else
  138. /* Run the below code for Cortex-M0 */
  139. float32_t squareOfSum; /* Square of Sum */
  140. if(blockSize == 1)
  141. {
  142. *pResult = 0;
  143. return;
  144. }
  145. /* Loop over blockSize number of values */
  146. blkCnt = blockSize;
  147. while(blkCnt > 0u)
  148. {
  149. /* C = (A[0] * A[0] + A[1] * A[1] + ... + A[blockSize-1] * A[blockSize-1]) */
  150. /* Compute Sum of squares of the input samples
  151. * and then store the result in a temporary variable, sumOfSquares. */
  152. in = *pSrc++;
  153. sumOfSquares += in * in;
  154. /* C = (A[0] + A[1] + ... + A[blockSize-1]) */
  155. /* Compute Sum of the input samples
  156. * and then store the result in a temporary variable, sum. */
  157. sum += in;
  158. /* Decrement the loop counter */
  159. blkCnt--;
  160. }
  161. /* Compute the square of sum */
  162. squareOfSum = ((sum * sum) / (float32_t) blockSize);
  163. /* Compute the variance */
  164. *pResult = ((sumOfSquares - squareOfSum) / (float32_t) (blockSize - 1.0f));
  165. #endif /* #ifndef ARM_MATH_CM0_FAMILY */
  166. }
  167. /**
  168. * @} end of variance group
  169. */