diff options
Diffstat (limited to 'docs/DSP/html/group__CmplxMatrixMult.html')
| -rw-r--r-- | docs/DSP/html/group__CmplxMatrixMult.html | 63 |
1 files changed, 35 insertions, 28 deletions
diff --git a/docs/DSP/html/group__CmplxMatrixMult.html b/docs/DSP/html/group__CmplxMatrixMult.html index 8e11533..e7a253b 100644 --- a/docs/DSP/html/group__CmplxMatrixMult.html +++ b/docs/DSP/html/group__CmplxMatrixMult.html @@ -32,7 +32,7 @@ <td id="projectlogo"><img alt="Logo" src="CMSIS_Logo_Final.png"/></td> <td style="padding-left: 0.5em;"> <div id="projectname">CMSIS-DSP -  <span id="projectnumber">Version 1.5.2</span> +  <span id="projectnumber">Version 1.7.0</span> </div> <div id="projectbrief">CMSIS DSP Software Library</div> </td> @@ -127,7 +127,12 @@ Functions</h2></td></tr> <tr class="separator:gaaf3c0b171ca8412c77bab9fa90804737"><td class="memSeparator" colspan="2"> </td></tr> </table> <a name="details" id="details"></a><h2 class="groupheader">Description</h2> -<p>Complex Matrix multiplication is only defined if the number of columns of the first matrix equals the number of rows of the second matrix. Multiplying an <code>M x N</code> matrix with an <code>N x P</code> matrix results in an <code>M x P</code> matrix. When matrix size checking is enabled, the functions check: (1) that the inner dimensions of <code>pSrcA</code> and <code>pSrcB</code> are equal; and (2) that the size of the output matrix equals the outer dimensions of <code>pSrcA</code> and <code>pSrcB</code>. </p> +<p>Complex Matrix multiplication is only defined if the number of columns of the first matrix equals the number of rows of the second matrix. Multiplying an <code>M x N</code> matrix with an <code>N x P</code> matrix results in an <code>M x P</code> matrix. </p> +<dl class="section user"><dt></dt><dd>When matrix size checking is enabled, the functions check:<ul> +<li>that the inner dimensions of <code>pSrcA</code> and <code>pSrcB</code> are equal;</li> +<li>that the size of the output matrix equals the outer dimensions of <code>pSrcA</code> and <code>pSrcB</code>. </li> +</ul> +</dd></dl> <h2 class="groupheader">Function Documentation</h2> <a class="anchor" id="ga1adb839ac84445b8c2f04efa43faef35"></a> <div class="memitem"> @@ -161,15 +166,17 @@ Functions</h2></td></tr> <p>Floating-point, complex, matrix multiplication.</p> <dl class="params"><dt>Parameters</dt><dd> <table class="params"> - <tr><td class="paramdir">[in]</td><td class="paramname">*pSrcA</td><td>points to the first input complex matrix structure </td></tr> - <tr><td class="paramdir">[in]</td><td class="paramname">*pSrcB</td><td>points to the second input complex matrix structure </td></tr> - <tr><td class="paramdir">[out]</td><td class="paramname">*pDst</td><td>points to output complex matrix structure </td></tr> + <tr><td class="paramdir">[in]</td><td class="paramname">pSrcA</td><td>points to first input complex matrix structure </td></tr> + <tr><td class="paramdir">[in]</td><td class="paramname">pSrcB</td><td>points to second input complex matrix structure </td></tr> + <tr><td class="paramdir">[out]</td><td class="paramname">pDst</td><td>points to output complex matrix structure </td></tr> </table> </dd> </dl> -<dl class="section return"><dt>Returns</dt><dd>The function returns either <code>ARM_MATH_SIZE_MISMATCH</code> or <code>ARM_MATH_SUCCESS</code> based on the outcome of size checking. </dd></dl> - -<p>References <a class="el" href="arm__math_8h.html#a5e459c6409dfcd2927bb8a57491d7cf6a7071b92f1f6bc3c5c312a237ea91105b">ARM_MATH_SIZE_MISMATCH</a>, <a class="el" href="arm__math_8h.html#a5e459c6409dfcd2927bb8a57491d7cf6a9f8b2a10bd827fb4600e77d455902eb0">ARM_MATH_SUCCESS</a>, <a class="el" href="structarm__matrix__instance__f32.html#acdd1fb73734df68b89565c54f1dd8ae2">arm_matrix_instance_f32::numCols</a>, <a class="el" href="structarm__matrix__instance__f32.html#a23f4e34d70a82c9cad7612add5640b7b">arm_matrix_instance_f32::numRows</a>, <a class="el" href="structarm__matrix__instance__f32.html#af3917c032600a9dfd5ed4a96f074910a">arm_matrix_instance_f32::pData</a>, and <a class="el" href="arm__dotproduct__example__f32_8c.html#a88ccb294236ab22b00310c47164c53c3">status</a>.</p> +<dl class="section return"><dt>Returns</dt><dd>execution status<ul> +<li><a class="el" href="arm__math_8h.html#a5e459c6409dfcd2927bb8a57491d7cf6a9f8b2a10bd827fb4600e77d455902eb0">ARM_MATH_SUCCESS</a> : Operation successful</li> +<li><a class="el" href="arm__math_8h.html#a5e459c6409dfcd2927bb8a57491d7cf6a7071b92f1f6bc3c5c312a237ea91105b">ARM_MATH_SIZE_MISMATCH</a> : Matrix size check failed </li> +</ul> +</dd></dl> </div> </div> @@ -211,21 +218,20 @@ Functions</h2></td></tr> <p>Q15, complex, matrix multiplication.</p> <dl class="params"><dt>Parameters</dt><dd> <table class="params"> - <tr><td class="paramdir">[in]</td><td class="paramname">*pSrcA</td><td>points to the first input complex matrix structure </td></tr> - <tr><td class="paramdir">[in]</td><td class="paramname">*pSrcB</td><td>points to the second input complex matrix structure </td></tr> - <tr><td class="paramdir">[out]</td><td class="paramname">*pDst</td><td>points to output complex matrix structure </td></tr> - <tr><td class="paramdir">[in]</td><td class="paramname">*pScratch</td><td>points to the array for storing intermediate results </td></tr> + <tr><td class="paramdir">[in]</td><td class="paramname">pSrcA</td><td>points to first input complex matrix structure </td></tr> + <tr><td class="paramdir">[in]</td><td class="paramname">pSrcB</td><td>points to second input complex matrix structure </td></tr> + <tr><td class="paramdir">[out]</td><td class="paramname">pDst</td><td>points to output complex matrix structure </td></tr> + <tr><td class="paramdir">[in]</td><td class="paramname">pScratch</td><td>points to an array for storing intermediate results </td></tr> </table> </dd> </dl> -<dl class="section return"><dt>Returns</dt><dd>The function returns either <code>ARM_MATH_SIZE_MISMATCH</code> or <code>ARM_MATH_SUCCESS</code> based on the outcome of size checking.</dd></dl> +<dl class="section return"><dt>Returns</dt><dd>execution status<ul> +<li><a class="el" href="arm__math_8h.html#a5e459c6409dfcd2927bb8a57491d7cf6a9f8b2a10bd827fb4600e77d455902eb0">ARM_MATH_SUCCESS</a> : Operation successful</li> +<li><a class="el" href="arm__math_8h.html#a5e459c6409dfcd2927bb8a57491d7cf6a7071b92f1f6bc3c5c312a237ea91105b">ARM_MATH_SIZE_MISMATCH</a> : Matrix size check failed</li> +</ul> +</dd></dl> <dl class="section user"><dt>Conditions for optimum performance</dt><dd>Input, output and state buffers should be aligned by 32-bit</dd></dl> -<dl class="section user"><dt>Restrictions</dt><dd>If the silicon does not support unaligned memory access enable the macro UNALIGNED_SUPPORT_DISABLE In this case input, output, scratch buffers should be aligned by 32-bit</dd></dl> -<p><b>Scaling and Overflow Behavior:</b></p> -<dl class="section user"><dt></dt><dd>The function is implemented using a 64-bit internal accumulator. The inputs to the multiplications are in 1.15 format and multiplications yield a 2.30 result. The 2.30 intermediate results are accumulated in a 64-bit accumulator in 34.30 format. This approach provides 33 guard bits and there is no risk of overflow. The 34.30 result is then truncated to 34.15 format by discarding the low 15 bits and then saturated to 1.15 format.</dd></dl> -<dl class="section user"><dt></dt><dd>Refer to <code><a class="el" href="group__MatrixMult.html#ga08f37d93a5bfef0c5000dc5e0a411f93" title="Q15 matrix multiplication (fast variant) for Cortex-M3 and Cortex-M4. ">arm_mat_mult_fast_q15()</a></code> for a faster but less precise version of this function. </dd></dl> - -<p>References <a class="el" href="arm__math_8h.html#a9de2e0a5785be82866bcb96012282248">__SIMD32</a>, <a class="el" href="arm__math_8h.html#a9c58d7b4648a973d91e27544bc190e60">__SMUADX()</a>, <a class="el" href="arm__math_8h.html#a6b293f62ad7eb225fa9ce0f39458eba3">__SMUSD()</a>, <a class="el" href="arm__math_8h.html#a5e459c6409dfcd2927bb8a57491d7cf6a7071b92f1f6bc3c5c312a237ea91105b">ARM_MATH_SIZE_MISMATCH</a>, <a class="el" href="arm__math_8h.html#a5e459c6409dfcd2927bb8a57491d7cf6a9f8b2a10bd827fb4600e77d455902eb0">ARM_MATH_SUCCESS</a>, <a class="el" href="structarm__matrix__instance__q15.html#acbbce67ba058d8e1c867c71d57288c97">arm_matrix_instance_q15::numCols</a>, <a class="el" href="structarm__matrix__instance__q15.html#a9bac6ed54be287c4d4f01a1a28be65f5">arm_matrix_instance_q15::numRows</a>, <a class="el" href="structarm__matrix__instance__q15.html#a6da33a5553e634787d0f515cf8d724af">arm_matrix_instance_q15::pData</a>, and <a class="el" href="arm__dotproduct__example__f32_8c.html#a88ccb294236ab22b00310c47164c53c3">status</a>.</p> +<dl class="section user"><dt>Scaling and Overflow Behavior</dt><dd>The function is implemented using an internal 64-bit accumulator. The inputs to the multiplications are in 1.15 format and multiplications yield a 2.30 result. The 2.30 intermediate results are accumulated in a 64-bit accumulator in 34.30 format. This approach provides 33 guard bits and there is no risk of overflow. The 34.30 result is then truncated to 34.15 format by discarding the low 15 bits and then saturated to 1.15 format. </dd></dl> </div> </div> @@ -261,17 +267,18 @@ Functions</h2></td></tr> <p>Q31, complex, matrix multiplication.</p> <dl class="params"><dt>Parameters</dt><dd> <table class="params"> - <tr><td class="paramdir">[in]</td><td class="paramname">*pSrcA</td><td>points to the first input complex matrix structure </td></tr> - <tr><td class="paramdir">[in]</td><td class="paramname">*pSrcB</td><td>points to the second input complex matrix structure </td></tr> - <tr><td class="paramdir">[out]</td><td class="paramname">*pDst</td><td>points to output complex matrix structure </td></tr> + <tr><td class="paramdir">[in]</td><td class="paramname">pSrcA</td><td>points to first input complex matrix structure </td></tr> + <tr><td class="paramdir">[in]</td><td class="paramname">pSrcB</td><td>points to second input complex matrix structure </td></tr> + <tr><td class="paramdir">[out]</td><td class="paramname">pDst</td><td>points to output complex matrix structure </td></tr> </table> </dd> </dl> -<dl class="section return"><dt>Returns</dt><dd>The function returns either <code>ARM_MATH_SIZE_MISMATCH</code> or <code>ARM_MATH_SUCCESS</code> based on the outcome of size checking.</dd></dl> -<p><b>Scaling and Overflow Behavior:</b></p> -<dl class="section user"><dt></dt><dd>The function is implemented using an internal 64-bit accumulator. The accumulator has a 2.62 format and maintains full precision of the intermediate multiplication results but provides only a single guard bit. There is no saturation on intermediate additions. Thus, if the accumulator overflows it wraps around and distorts the result. The input signals should be scaled down to avoid intermediate overflows. The input is thus scaled down by log2(numColsA) bits to avoid overflows, as a total of numColsA additions are performed internally. The 2.62 accumulator is right shifted by 31 bits and saturated to 1.31 format to yield the final result. </dd></dl> - -<p>References <a class="el" href="arm__math_8h.html#a5e459c6409dfcd2927bb8a57491d7cf6a7071b92f1f6bc3c5c312a237ea91105b">ARM_MATH_SIZE_MISMATCH</a>, <a class="el" href="arm__math_8h.html#a5e459c6409dfcd2927bb8a57491d7cf6a9f8b2a10bd827fb4600e77d455902eb0">ARM_MATH_SUCCESS</a>, <a class="el" href="arm__math_8h.html#a9a94b78de0f083bd8fadf3c3abe40431">clip_q63_to_q31()</a>, <a class="el" href="structarm__matrix__instance__q31.html#abd161da7614eda927157f18b698074b1">arm_matrix_instance_q31::numCols</a>, <a class="el" href="structarm__matrix__instance__q31.html#a63bacac158a821c8cfc06088d251598c">arm_matrix_instance_q31::numRows</a>, <a class="el" href="structarm__matrix__instance__q31.html#a09a64267c0579fef086efc9059741e56">arm_matrix_instance_q31::pData</a>, and <a class="el" href="arm__dotproduct__example__f32_8c.html#a88ccb294236ab22b00310c47164c53c3">status</a>.</p> +<dl class="section return"><dt>Returns</dt><dd>execution status<ul> +<li><a class="el" href="arm__math_8h.html#a5e459c6409dfcd2927bb8a57491d7cf6a9f8b2a10bd827fb4600e77d455902eb0">ARM_MATH_SUCCESS</a> : Operation successful</li> +<li><a class="el" href="arm__math_8h.html#a5e459c6409dfcd2927bb8a57491d7cf6a7071b92f1f6bc3c5c312a237ea91105b">ARM_MATH_SIZE_MISMATCH</a> : Matrix size check failed</li> +</ul> +</dd></dl> +<dl class="section user"><dt>Scaling and Overflow Behavior</dt><dd>The function is implemented using an internal 64-bit accumulator. The accumulator has a 2.62 format and maintains full precision of the intermediate multiplication results but provides only a single guard bit. There is no saturation on intermediate additions. Thus, if the accumulator overflows it wraps around and distorts the result. The input signals should be scaled down to avoid intermediate overflows. The input is thus scaled down by log2(numColsA) bits to avoid overflows, as a total of numColsA additions are performed internally. The 2.62 accumulator is right shifted by 31 bits and saturated to 1.31 format to yield the final result. </dd></dl> </div> </div> @@ -280,7 +287,7 @@ Functions</h2></td></tr> <!-- start footer part --> <div id="nav-path" class="navpath"><!-- id is needed for treeview function! --> <ul> - <li class="footer">Generated on Wed Aug 1 2018 17:12:22 for CMSIS-DSP by Arm Ltd. All rights reserved. + <li class="footer">Generated on Wed Jul 10 2019 15:20:40 for CMSIS-DSP Version 1.7.0 by Arm Ltd. All rights reserved. <!-- <a href="http://www.doxygen.org/index.html"> <img class="footer" src="doxygen.png" alt="doxygen"/></a> 1.8.6 |