arm_matrix_example_f32.c
1 |
/* ----------------------------------------------------------------------
|
---|---|
2 |
* Copyright (C) 2010 ARM Limited. All rights reserved.
|
3 |
*
|
4 |
* $Date: 29. November 2010
|
5 |
* $Revision: V1.0.3
|
6 |
*
|
7 |
* Project: CMSIS DSP Library
|
8 |
* Title: arm_matrix_example_f32.c
|
9 |
*
|
10 |
* Description: Example code demonstrating least square fit to data
|
11 |
* using matrix functions
|
12 |
*
|
13 |
* Target Processor: Cortex-M4/Cortex-M3
|
14 |
*
|
15 |
*
|
16 |
* Version 1.0.3 2010/11/29
|
17 |
* Re-organized the CMSIS folders and updated documentation.
|
18 |
*
|
19 |
* Version 1.0.1 2010/10/05 KK
|
20 |
* Production release and review comments incorporated.
|
21 |
*
|
22 |
* Version 1.0.0 2010/09/20 KK
|
23 |
* Production release and review comments incorporated.
|
24 |
* ------------------------------------------------------------------- */
|
25 |
|
26 |
/**
|
27 |
* @ingroup groupExamples
|
28 |
*/
|
29 |
|
30 |
/**
|
31 |
* @defgroup MatrixExample Matrix Example
|
32 |
*
|
33 |
* \par Description:
|
34 |
* \par
|
35 |
* Demonstrates the use of Matrix Transpose, Matrix Muliplication, and Matrix Inverse
|
36 |
* functions to apply least squares fitting to input data. Least squares fitting is
|
37 |
* the procedure for finding the best-fitting curve that minimizes the sum of the
|
38 |
* squares of the offsets (least square error) from a given set of data.
|
39 |
*
|
40 |
* \par Algorithm:
|
41 |
* \par
|
42 |
* The linear combination of parameters considered is as follows:
|
43 |
* \par
|
44 |
* <code>A * X = B</code>, where \c X is the unknown value and can be estimated
|
45 |
* from \c A & \c B.
|
46 |
* \par
|
47 |
* The least squares estimate \c X is given by the following equation:
|
48 |
* \par
|
49 |
* <code>X = Inverse(A<sup>T</sup> * A) * A<sup>T</sup> * B</code>
|
50 |
*
|
51 |
* \par Block Diagram:
|
52 |
* \par
|
53 |
* \image html matrixExample.gif
|
54 |
*
|
55 |
* \par Variables Description:
|
56 |
* \par
|
57 |
* \li \c A_f32 input matrix in the linear combination equation
|
58 |
* \li \c B_f32 output matrix in the linear combination equation
|
59 |
* \li \c X_f32 unknown matrix estimated using \c A_f32 & \c B_f32 matrices
|
60 |
*
|
61 |
* \par CMSIS DSP Software Library Functions Used:
|
62 |
* \par
|
63 |
* - arm_mat_init_f32()
|
64 |
* - arm_mat_trans_f32()
|
65 |
* - arm_mat_mult_f32()
|
66 |
* - arm_mat_inverse_f32()
|
67 |
*
|
68 |
* <b> Refer </b>
|
69 |
* \link arm_matrix_example_f32.c \endlink
|
70 |
*
|
71 |
*/
|
72 |
|
73 |
|
74 |
/** \example arm_matrix_example_f32.c
|
75 |
*/
|
76 |
|
77 |
#include "arm_math.h" |
78 |
#include "math_helper.h" |
79 |
|
80 |
#define SNR_THRESHOLD 90 |
81 |
|
82 |
/* --------------------------------------------------------------------------------
|
83 |
* Test input data(Cycles) taken from FIR Q15 module for differant cases of blockSize
|
84 |
* and tapSize
|
85 |
* --------------------------------------------------------------------------------- */
|
86 |
|
87 |
const float32_t B_f32[4] = |
88 |
{ |
89 |
782.0, 7577.0, 470.0, 4505.0 |
90 |
}; |
91 |
|
92 |
/* --------------------------------------------------------------------------------
|
93 |
* Formula to fit is C1 + C2 * numTaps + C3 * blockSize + C4 * numTaps * blockSize
|
94 |
* -------------------------------------------------------------------------------- */
|
95 |
|
96 |
const float32_t A_f32[16] = |
97 |
{ |
98 |
/* Const, numTaps, blockSize, numTaps*blockSize */
|
99 |
1.0, 32.0, 4.0, 128.0, |
100 |
1.0, 32.0, 64.0, 2048.0, |
101 |
1.0, 16.0, 4.0, 64.0, |
102 |
1.0, 16.0, 64.0, 1024.0, |
103 |
}; |
104 |
|
105 |
|
106 |
/* ----------------------------------------------------------------------
|
107 |
* Temporary buffers for storing intermediate values
|
108 |
* ------------------------------------------------------------------- */
|
109 |
/* Transpose of A Buffer */
|
110 |
float32_t AT_f32[16];
|
111 |
/* (Transpose of A * A) Buffer */
|
112 |
float32_t ATMA_f32[16];
|
113 |
/* Inverse(Transpose of A * A) Buffer */
|
114 |
float32_t ATMAI_f32[16];
|
115 |
/* Test Output Buffer */
|
116 |
float32_t X_f32[4];
|
117 |
|
118 |
/* ----------------------------------------------------------------------
|
119 |
* Reference ouput buffer C1, C2, C3 and C4 taken from MATLAB
|
120 |
* ------------------------------------------------------------------- */
|
121 |
const float32_t xRef_f32[4] = {73.0, 8.0, 21.25, 2.875}; |
122 |
|
123 |
float32_t snr; |
124 |
|
125 |
|
126 |
/* ----------------------------------------------------------------------
|
127 |
* Max magnitude FFT Bin test
|
128 |
* ------------------------------------------------------------------- */
|
129 |
|
130 |
int32_t main(void)
|
131 |
{ |
132 |
|
133 |
arm_matrix_instance_f32 A; /* Matrix A Instance */
|
134 |
arm_matrix_instance_f32 AT; /* Matrix AT(A transpose) instance */
|
135 |
arm_matrix_instance_f32 ATMA; /* Matrix ATMA( AT multiply with A) instance */
|
136 |
arm_matrix_instance_f32 ATMAI; /* Matrix ATMAI(Inverse of ATMA) instance */
|
137 |
arm_matrix_instance_f32 B; /* Matrix B instance */
|
138 |
arm_matrix_instance_f32 X; /* Matrix X(Unknown Matrix) instance */
|
139 |
|
140 |
uint32_t srcRows, srcColumns; /* Temporary variables */
|
141 |
arm_status status; |
142 |
|
143 |
/* Initialise A Matrix Instance with numRows, numCols and data array(A_f32) */
|
144 |
srcRows = 4;
|
145 |
srcColumns = 4;
|
146 |
arm_mat_init_f32(&A, srcRows, srcColumns, (float32_t *)A_f32); |
147 |
|
148 |
/* Initialise Matrix Instance AT with numRows, numCols and data array(AT_f32) */
|
149 |
srcRows = 4;
|
150 |
srcColumns = 4;
|
151 |
arm_mat_init_f32(&AT, srcRows, srcColumns, AT_f32); |
152 |
|
153 |
/* calculation of A transpose */
|
154 |
status = arm_mat_trans_f32(&A, &AT); |
155 |
|
156 |
|
157 |
/* Initialise ATMA Matrix Instance with numRows, numCols and data array(ATMA_f32) */
|
158 |
srcRows = 4;
|
159 |
srcColumns = 4;
|
160 |
arm_mat_init_f32(&ATMA, srcRows, srcColumns, ATMA_f32); |
161 |
|
162 |
/* calculation of AT Multiply with A */
|
163 |
status = arm_mat_mult_f32(&AT, &A, &ATMA); |
164 |
|
165 |
/* Initialise ATMAI Matrix Instance with numRows, numCols and data array(ATMAI_f32) */
|
166 |
srcRows = 4;
|
167 |
srcColumns = 4;
|
168 |
arm_mat_init_f32(&ATMAI, srcRows, srcColumns, ATMAI_f32); |
169 |
|
170 |
/* calculation of Inverse((Transpose(A) * A) */
|
171 |
status = arm_mat_inverse_f32(&ATMA, &ATMAI); |
172 |
|
173 |
/* calculation of (Inverse((Transpose(A) * A)) * Transpose(A)) */
|
174 |
status = arm_mat_mult_f32(&ATMAI, &AT, &ATMA); |
175 |
|
176 |
/* Initialise B Matrix Instance with numRows, numCols and data array(B_f32) */
|
177 |
srcRows = 4;
|
178 |
srcColumns = 1;
|
179 |
arm_mat_init_f32(&B, srcRows, srcColumns, (float32_t *)B_f32); |
180 |
|
181 |
/* Initialise X Matrix Instance with numRows, numCols and data array(X_f32) */
|
182 |
srcRows = 4;
|
183 |
srcColumns = 1;
|
184 |
arm_mat_init_f32(&X, srcRows, srcColumns, X_f32); |
185 |
|
186 |
/* calculation ((Inverse((Transpose(A) * A)) * Transpose(A)) * B) */
|
187 |
status = arm_mat_mult_f32(&ATMA, &B, &X); |
188 |
|
189 |
/* Comparison of reference with test output */
|
190 |
snr = arm_snr_f32((float32_t *)xRef_f32, X_f32, 4);
|
191 |
|
192 |
/*------------------------------------------------------------------------------
|
193 |
* Initialise status depending on SNR calculations
|
194 |
*------------------------------------------------------------------------------*/
|
195 |
if( snr > SNR_THRESHOLD)
|
196 |
{ |
197 |
status = ARM_MATH_SUCCESS; |
198 |
} |
199 |
else
|
200 |
{ |
201 |
status = ARM_MATH_TEST_FAILURE; |
202 |
} |
203 |
|
204 |
|
205 |
/* ----------------------------------------------------------------------
|
206 |
** Loop here if the signals fail the PASS check.
|
207 |
** This denotes a test failure
|
208 |
** ------------------------------------------------------------------- */
|
209 |
if( status != ARM_MATH_SUCCESS)
|
210 |
{ |
211 |
while(1); |
212 |
} |
213 |
|
214 |
while(1); /* main function does not return */ |
215 |
} |
216 |
|
217 |
/** \endlink */
|
218 |
|