| /cmsis-dsp-latest/dsppp/Include/dsppp/Helium/ |
| D | float.hpp | 87 * @return Value with scalar datatype
124 * @return Vector initialized with constant in each lane
132 * @brief Vector constant with tail
137 * @return Vector initialized with constant in some lanes
159 * @brief Vector negate with tail
212 * @brief Vector + Vector with tail
218 * @return a + b with tail predicate
228 * @brief Vector + scalar with tail
234 * @return a + b with tail predicate
243 * @brief Scalar + vector with tail predicate
[all …]
|
| D | half.hpp | 77 * @return Lane value with current datatype
122 * @brief Vector of const with tail predicate
127 * @return The initialized vector with const and predicate
148 * @brief Vector negate with tail predicate
153 * @return Negate of vector with tail predicate
210 * @brief Vector + Vector with tail predicate
216 * @return a + b with tail predicate
226 * @brief Vector + Scalar with tail predicate
232 * @return a + b with tail predicate
242 * @brief Scalar + Vector with tail predicate
[all …]
|
| /cmsis-dsp-latest/ |
| D | README.md | 37 Kernels are provided with several datatypes : f64, f32, f16, q31, q15, q7.
41 A [PythonWrapper](https://pypi.org/project/cmsisdsp/) is also available and can be installed with:
45 With this wrapper you can design your algorithm in Python using an API as close as possible to the …
60 * [How to build with MDK or Open CMSIS-Pack](#how-to-build-with-mdk-or-open-cmsis-pack)
61 * [How to build with Make](#how-to-build-with-make)
62 * [How to build with cmake](#how-to-build-with-cmake)
63 * [How to build with any other build system](#how-to-build-with-any-other-build-system)
72 CMSIS-DSP is used when you need performance. As consequence CMSIS-DSP should be compiled with the o…
82 …with Helium support, it will be automatically detected by CMSIS-DSP. For Neon, it is not the case …
84 * `-DLOOPUNROLL=ON` can also be used when compiling with cmake
[all …]
|
| D | PythonWrapper_README.md | 3 …en source [CMSIS-DSP](https://github.com/ARM-software/CMSIS-DSP) and it is compatible with `NumPy`.
11 A tutorial is also available but with less details than this README:
20 The building of this package has been tested on Windows with the Python install from python.org and…
22 It has also been tested with `cygwin`. In that case, `python-devel` must be installed too. On Mac, …
65 …folders in `PythonWrapper` are giving some examples of the options to use with the `cmake` command…
101 ## Functions with no instance arguments
103 You can use a [CMSIS-DSP](https://github.com/ARM-software/CMSIS-DSP) function with numpy arrays:
107 The function can also be called more simply with
113 ## Functions with instance arguments
125 …I and in the final C implementation you'll need to allocate a state array with the right dimension.
[all …]
|
| D | LICENSE | 17 control with that entity. For the purposes of this definition,
81 with the Work to which such Contribution(s) was submitted. If You
90 Work or Derivative Works thereof in any medium, with or without
113 documentation, if provided along with the Derivative Works; or,
127 reproduction, and distribution of the Work otherwise complies with
136 with Licensor regarding such Contributions.
151 risks associated with Your exercise of permissions under this License.
168 or other liability obligations and/or rights consistent with this
181 boilerplate notice, with the fields enclosed by brackets "{}"
182 replaced with your own identifying information. (Don't include
[all …]
|
| /cmsis-dsp-latest/Documentation/Doxygen/src/ |
| D | memory_static_dynamic.md | 5 * `Vector<T>` with a dimension known at runtime
6 * `Vector<T,NB>` with a dimension known at build time
10 This naming "static" / "dynamic" is referring to the dimension. With "dynamic" vectors the same cod…
12 With "static" vectors : the length is fixed at build time and will never change at runtime.
20 With static objects it is also possible to use different memory allocator with better performances …
22 But, with static objects, objects of different dimension are considered as different types. The com…
28 With dynamic objects, the dimension is know at runtime. So object of different dimensions have the …
32 With vector instructions one can use scatter / gather instructions and they require a stride. But t…
34 Finally, with dynamic object, memory allocation can be an issue. You can mitigate the problem by re…
|
| D | template.md | 12 …ons using C preprocessor. But we would still have two different functions with different names at …
14 With C++ templates, we can achieve the same result in a better way since the C++ compiler will chec…
16 With C++ template, we could have a *generic* function `arm_add` taking as argument a pointer `T *pS…
18 When the function is used with a `float32_t *`, the compiler would generate code for a function usi…
20 And if the function is used with a `float64_t *`, the compiler would generate code for a function u…
24 …ate an implementation because the type variable `T` is replaced by a type with no addition operato…
50 In the first case, the length is an implicit parameter with a default value and it is equivalent to…
60 Now you can look at an @ref dsppp_vector_example "example with vector operations" showing how to us…
|
| D | introduction.md | 13 with CMSIS-DSP, you would write:
22 There are several limitations with this way of writing the code:
32 With this new C++ template library, you can write:
39 The code generated by this line computes the dot product in one pass with all the operators (`+`, `…
52 With the C++ library, it can be written as:
59 Here again : all the vector operations (`+`,`*`) are done in one pass with one loop. There is no mo…
63 You can also jump directly to an @ref dsppp_vector_example "example with vector operations".
|
| D | vectorop.md | 44 …2_t`. This creation is requiring some memory allocation and by default it is done with a `malloc`.
52 If you don't know the dimension at build time, you can use a different type of vector with:
58 For the trade-off between vector with build time dimension or runtime dimension please see the sect…
79 …e that the computation can be parametrized with template arguments so the same computation could b…
87 The operators `+`, `*` are computed in one pass with one loop : we have loop fusion and instead of …
89 To understand fusion and how to extend it with new operators, see section @ref dsppp_fusion .
|
| D | vector.md | 3 The use of vectors has been explained in @ref dsppp_vector_example "example with vector operations"…
21 Example with `Q15` is very similar:
58 One can define a `VectorView` with:
66 You can then operate with this virtual vector:
88 It is advised to always use the `copy` operator (even with normal vectors).
96 This line sets the odd elements of the vector to `0.0f`. It is creating a virtual vector with strid…
|
| D | mainpage.md | 35 …with a Python API as close as possible to the C one. It can be used to start developing and testin…
63 The library ships with a number of examples which demonstrate how to use the library functions. Ple…
67 The library is now tested on Fast Models building with cmake. Core M0, M4, M7, M33, M55 are tested.
78 �� Documentation | Folder with this CMSIS-DSP documenation
130 …- With Helium or Neon, disable the use of vectorized code with C intrinsics and use pure C instead…
140 Previous versions were using lots of compilation flags to control code size. It was enabled with `A…
162 All algorithms are compared with a double precision reference and the different versions (for diffe…
|
| /cmsis-dsp-latest/Testing/ |
| D | README.md | 7 Here is a quick summary of how to get started with the framework the first time the repository is c…
121 A test suite is a set of tests packaged with some data.
151 A test description file is defined with a specific syntax to support R1 to R8.
188 A function is described with some text and followed by the name of the function in the C++ class.
214 The file is in the folder defined with the folder properties of the group / suites.
220 A benchmark will often have to be run with different lengths for the input.
224 In the benchmark results, we may want to generate a CSV (or any other format) with different column…
255 Those parameters values, when specified with a file, are described with:
262 When the parameter is specified with a generator then the syntax is :
272 To use parameters with a function the syntax is:
[all …]
|
| /cmsis-dsp-latest/PythonWrapper/docs/source/ |
| D | api.rst | 49 You can use a CMSIS-DSP function with numpy arrays: ::
55 Functions with instance arguments
93 …I and in the final C implementation you'll need to allocate a state array with the right dimension.
97 … other arguments. This choice was made to make it a bit easier the use of numpy array with the API.
126 Then, you can filter with CMSIS-DSP::
136 If you want to compare with scipy it is easy but warning : coefficients for the filter are in oppos…
144 The CMSIS-DSP cfft is requiring complex signals with a specific layout in memory.
154 …FFT will not be containing complex Python scalars. It must be converted back with a function like::
161 Then, you create the FFT instance with::
165 You initialize the instance with the init function ::
[all …]
|
| /cmsis-dsp-latest/dsppp/Include/dsppp/ |
| D | matrix_view.hpp | 203 /** @brief Elementwise multiply matrix view with expression
216 /** @brief Elementwise multiply matrix view with matrix view
258 /** @brief Create a row view with stride 1
269 /** @brief Create a row view with stride 1
281 /** @brief Create a constant row view with stride 1
292 /** @brief Create a constant row view with stride 1
382 * @brief %Vector store at a given row,column position with predicated tail
387 * @param val Vector value to write at index i with tail predication
392 * with predication
403 * @brief %Vector operation at a given row,column position with predicated tail
[all …]
|
| /cmsis-dsp-latest/Testing/Source/Tests/ |
| D | BIQUADF16.cpp | 9 Reference patterns are generated with
38 Python script is generating different tests with in test_biquad_cascade_df1_ref()
50 The filter is initialized with the coefs, blockSize and numTaps. in test_biquad_cascade_df1_ref()
108 Python script is generating different tests with in test_biquad_cascade_df2T_ref()
119 The filter is initialized with the coefs, blockSize and numTaps. in test_biquad_cascade_df2T_ref()
179 Python script is generating different tests with in test_biquad_cascade_df1_rand()
194 The filter is initialized with the coefs, blockSize and numTaps. in test_biquad_cascade_df1_rand()
254 Python script is generating different tests with in test_biquad_cascade_df2T_rand()
270 The filter is initialized with the coefs, blockSize and numTaps. in test_biquad_cascade_df2T_rand()
326 Python script is generating different tests with in test_biquad_cascade_stereo_df2T_rand()
[all …]
|
| D | BIQUADF32.cpp | 9 Reference patterns are generated with
38 Python script is generating different tests with in test_biquad_cascade_df1_ref()
49 The filter is initialized with the coefs, blockSize and numTaps. in test_biquad_cascade_df1_ref()
107 Python script is generating different tests with in test_biquad_cascade_df2T_ref()
118 The filter is initialized with the coefs, blockSize and numTaps. in test_biquad_cascade_df2T_ref()
189 Python script is generating different tests with in test_biquad_cascade_df1_rand()
204 The filter is initialized with the coefs, blockSize and numTaps. in test_biquad_cascade_df1_rand()
268 Python script is generating different tests with in test_biquad_cascade_df2T_rand()
284 The filter is initialized with the coefs, blockSize and numTaps. in test_biquad_cascade_df2T_rand()
352 Python script is generating different tests with in test_biquad_cascade_stereo_df2T_rand()
[all …]
|
| /cmsis-dsp-latest/.github/workflows/ |
| D | runcpptest.yaml | 18 # GCC currently has too many problems with the C++ API
29 with:
41 with:
49 with:
58 with:
98 with:
|
| /cmsis-dsp-latest/Examples/ARM/arm_svm_example/ |
| D | arm_svm_example_f32.c | 14 * Redistribution and use in source and binary forms, with or without
21 * the documentation and/or other materials provided with the
50 * about classical ML with CMSIS-DSP and python scikit-learn:
51 …e-learning-on-arm/developer-material/how-to-guides/implement-classical-ml-with-arm-cmsis-dsp-libra…
63 Those parameters can be generated with the python library scikit-learn.
92 Class A is identified with value 0.
93 Class B is identified with value 1.
96 with the Python code where different values could be used.
|
| /cmsis-dsp-latest/Source/TransformFunctions/ |
| D | arm_cfft_init_f64.c | 17 * not use this file except in compliance with the License.
76 @brief Initialization function for the cfft f64 function with 4096 samples
89 @brief Initialization function for the cfft f64 function with 2048 samples
103 @brief Initialization function for the cfft f64 function with 1024 samples
116 @brief Initialization function for the cfft f64 function with 512 samples
129 @brief Initialization function for the cfft f64 function with 256 samples
142 @brief Initialization function for the cfft f64 function with 128 samples
155 @brief Initialization function for the cfft f64 function with 64 samples
168 @brief Initialization function for the cfft f64 function with 32 samples
181 @brief Initialization function for the cfft f64 function with 16 samples
|
| D | arm_dct4_f32.c | 17 * not use this file except in compliance with the License.
48 computing the DFT of a real signal with a few additional operations.
52 …DCT4 is implemented using DCT2 as their implementations are similar except with some added pre-pro…
74 …T4 matrices become involutory (i.e. they are self-inverse) by multiplying with an overall scale fa…
144 * along with some pre-processing and post-processing. in arm_dct4_f32()
146 * (a) Pre-processing involves multiplying input with cos factor, in arm_dct4_f32()
159 * (d) Multiplying the output with the normalizing factor sqrt(2/N). in arm_dct4_f32()
163 /* Multiplying input with cos factor i.e. r(n) = 2 * x(n) * cos(pi*(2*n+1)/(4*n)) */ in arm_dct4_f32()
188 /* First part of the processing with loop unrolling. Compute 4 outputs at a time. in arm_dct4_f32()
220 /* Processing with loop unrolling 4 times as N is always multiple of 4. in arm_dct4_f32()
[all …]
|
| D | arm_cfft_init_f32.c | 17 * not use this file except in compliance with the License.
127 @brief Initialization function for the cfft f32 function with 4096 samples
141 @brief Initialization function for the cfft f32 function with 2048 samples
155 @brief Initialization function for the cfft f32 function with 1024 samples
169 @brief Initialization function for the cfft f32 function with 512 samples
183 @brief Initialization function for the cfft f32 function with 256 samples
197 @brief Initialization function for the cfft f32 function with 128 samples
211 @brief Initialization function for the cfft f32 function with 64 samples
225 @brief Initialization function for the cfft f32 function with 32 samples
239 @brief Initialization function for the cfft f32 function with 16 samples
|
| D | arm_cfft_init_f16.c | 17 * not use this file except in compliance with the License.
137 @brief Initialization function for the cfft f16 function with 4096 samples
150 @brief Initialization function for the cfft f16 function with 2048 samples
164 @brief Initialization function for the cfft f16 function with 1024 samples
178 @brief Initialization function for the cfft f16 function with 512 samples
192 @brief Initialization function for the cfft f16 function with 256 samples
206 @brief Initialization function for the cfft f16 function with 128 samples
220 @brief Initialization function for the cfft f16 function with 64 samples
234 @brief Initialization function for the cfft f16 function with 32 samples
248 @brief Initialization function for the cfft f16 function with 16 samples
|
| D | arm_dct4_q15.c | 17 * not use this file except in compliance with the License.
70 * along with some pre-processing and post-processing. in arm_dct4_q15()
72 * (a) Pre-processing involves multiplying input with cos factor, in arm_dct4_q15()
85 * (d) Multiplying the output with the normalizing factor sqrt(2/N). in arm_dct4_q15()
89 /* Multiplying input with cos factor i.e. r(n) = 2 * x(n) * cos(pi*(2*n+1)/(4*n)) */ in arm_dct4_q15()
114 /* First part of the processing with loop unrolling. Compute 4 outputs at a time. in arm_dct4_q15()
146 /* Processing with loop unrolling 4 times as N is always multiple of 4. in arm_dct4_q15()
168 * Step3: Multiply the FFT output with the weights. in arm_dct4_q15()
199 /* First part of the processing with loop unrolling. Compute 4 outputs at a time. in arm_dct4_q15()
245 /*------------ Normalizing the output by multiplying with the normalizing factor ----------*/ in arm_dct4_q15()
[all …]
|
| /cmsis-dsp-latest/Source/SupportFunctions/ |
| D | arm_quick_sort_f32.c | 17 * not use this file except in compliance with the License.
63 /* Compare left elements with pivot */ in arm_quick_sort_partition_f32()
69 /* Compare right elements with pivot */ in arm_quick_sort_partition_f32()
77 /* Compare left elements with pivot */ in arm_quick_sort_partition_f32()
83 /* Compare right elements with pivot */ in arm_quick_sort_partition_f32()
113 /* Iterate algorithm with two sub-arrays [first ... pivot] and [pivot+1 ... last] */ in arm_quick_sort_core_f32()
139 * is chosen, all the elements with values smaller than the
140 * pivot are moved before the pivot, while all elements with
|
| /cmsis-dsp-latest/Examples/ARM/arm_matrix_example/ |
| D | arm_matrix_example_f32.c | 15 * Redistribution and use in source and binary forms, with or without
22 * the documentation and/or other materials provided with the
151 arm_matrix_instance_f32 ATMA; /* Matrix ATMA( AT multiply with A) instance */ in main()
159 /* Initialise A Matrix Instance with numRows, numCols and data array(A_f32) */ in main()
164 /* Initialise Matrix Instance AT with numRows, numCols and data array(AT_f32) */ in main()
173 /* Initialise ATMA Matrix Instance with numRows, numCols and data array(ATMA_f32) */ in main()
178 /* calculation of AT Multiply with A */ in main()
181 /* Initialise ATMAI Matrix Instance with numRows, numCols and data array(ATMAI_f32) */ in main()
192 /* Initialise B Matrix Instance with numRows, numCols and data array(B_f32) */ in main()
197 /* Initialise X Matrix Instance with numRows, numCols and data array(X_f32) */ in main()
[all …]
|
