|
|
41e748c2f5
|
Made code to compile and probably work with 32-bits floating point. This requires quite some testing to be done
Building, testing and releasing LASP if it has a tag / Build-Test-Ubuntu (push) Failing after -4m54s
Building, testing and releasing LASP if it has a tag / Release-Ubuntu (push) Has been skipped
|
2024-06-03 17:28:51 +02:00 |
|
|
|
9617da3ad9
|
Updated libs. Some comments and tests on cmake with msys2.
continuous-integration/drone/push Build is failing
|
2023-05-25 11:41:16 -07:00 |
|
|
|
0ed005e35f
|
Cleanup of tests
continuous-integration/drone/push Build is failing
|
2023-01-12 20:35:00 +01:00 |
|
|
|
d1dea1483f
|
Updated tests
continuous-integration/drone/push Build was killed
|
2023-01-12 20:31:55 +01:00 |
|
|
|
0e31dbceac
|
Bugfix: FilterBankDesigner now always returns second order sections with shape (SOS_ORDER, 6), even for the high pass section. Expanded the third octave bank compliance test. Since it passed for all common sample rates, I removed the warning that the filters might be non-compliant for fs unequal to 48 kHz.
|
2023-01-09 11:59:48 +01:00 |
|
|
|
786afc3fac
|
FilterBankDesigner now handles sampling frequencies other than 48 kHz; added norm compliance test script. If part of the (1/3 or 1/1) octave band lies above the Nyquist frequency, a high pass filter is returned. If the whole band lies above the Nyquist frequency, a no pass filter is returned.
|
2023-01-06 09:44:34 +01:00 |
|
|
|
ec689621b5
|
Added Armadillo backend for Fft's. This one seems to work properly. Added some tests code for daq and StreamMgr in C++. Fixed a stray of a debug dmat somewhere in lasp_daqdata.
|
2022-10-12 15:02:42 +02:00 |
|
|
|
4764a52de8
|
Improved code for cubes. RtAps is not working. Don't know why. TimeBuffer code has better readability. Bugfix in output from Column to Numpy. GIL release for AvPowerSpectra::compute via Pybind.
|
2022-10-11 14:50:44 +02:00 |
|
|
|
ebe385b9db
|
Cleaned up example code
|
2022-10-10 19:29:24 +02:00 |
|
|
|
3481e4f9ba
|
Bugfix for OpenMP combinded with FFT. That one could not run in parallel in its current form.
|
2022-10-05 14:57:39 +02:00 |
|
|
|
12d6826140
|
Removed some compile warnings when compiling sub-projects. Removed warnings related to unknown pragmas in case of compiling without OpenMP support. Fixed some test problems
|
2022-10-05 13:39:45 +02:00 |
|
|
|
5f1a207104
|
Added Peak Programme Meter, added functionality to clone SeriesBiquads without copying state. Improved recording callback mechanism. Added reset() function for IndataHandlers, which send a pointer to DAQ instance.
|
2022-10-04 09:27:27 +02:00 |
|
|
|
bb26fc6bcc
|
Added suggested SLM down sampling factor, improved quite a lot of documentation. Measurement object can now work with old, as well as new measurement files.
|
2022-10-01 19:59:35 +02:00 |
|
|
|
b629edde4c
|
Updated pyproject and doxyfile, and readme
|
2022-09-22 10:11:47 +02:00 |
|
|
|
70891ceaf4
|
Switched to OpenMP for parallellizing for loops. Bugfixes in PowerSpectra() class. Added tests to check Parsevall hold. Improved lots of comments. Added module groups. Use CMake to specify code version number. Device info should be obtained from StreamMgr in Python code.
|
2022-09-03 20:59:14 +02:00 |
|
|
|
34239bfabf
|
Removed old testing code. Improved Doxyfile, added logo to Doxygen. Added pytest testing config.
|
2022-09-03 16:07:44 +02:00 |
|
|
|
c75f0dddc5
|
Split up indatahandler with python callback in part and part that calls the Python function. Threading is now handled using a thread pool. Some bugfixes
|
2022-08-14 21:00:22 +02:00 |
|
|
|
6a006e27f9
|
Output of a sine wave now works. Pfff
|
2022-07-20 14:58:48 +02:00 |
|
|
|
9983ca6096
|
Throughput of flag LASP_PARALLEL. If not set, some downstream code will be broken as JobQueues will not go empty for this case. For all code in the library, a serial implementation works off the not present job queue
|
2021-09-15 21:10:27 +02:00 |
|
|
|
b159f4d144
|
Major cleanup of build code (CMake related). First steps towards C++ and C-compiler safety. Also first steps made to get single precision back to working
|
2021-09-14 20:05:42 +02:00 |
|
|
|
7153096552
|
Fixed several bugs. Most serious one is a segfault for a stream without input.
|
2021-05-19 16:33:27 +02:00 |
|
|
|
fa32e9eac3
|
Bugfixes of Python/C++ interaction layer, first tests for DaqConfiguration code
|
2020-10-18 13:58:42 +02:00 |
|
|
|
f54ac216c1
|
Major cleanup of compilation warnings.
|
2020-10-16 20:37:02 +02:00 |
|
|
|
bc9639f10c
|
Inbetween state. Nothing is working
|
2020-10-14 09:42:19 +02:00 |
|
|
|
29662c82e3
|
Working C++ multiple API sit
|
2020-10-10 18:28:43 +02:00 |
|
|
|
a3963c4595
|
Working in CPP code, with multiple-api-api
|
2020-10-06 21:29:02 +02:00 |
|
|
|
a73ef3d7a8
|
First work on moving RtAudio Api back to C++.
|
2020-10-04 22:29:54 +02:00 |
|
|
|
4cd3f0bf9f
|
Added Second Order Sections filterbank. Works much better that Fir filterbank. Does not need decimation. Easy generation of filters with scipy.signal.butter.
|
2020-01-03 21:11:56 +01:00 |
|
|
|
195319ab29
|
Added the possibility to shift to different fft backend. Now set to fftw. Seems to work properly
|
2019-12-29 22:07:27 +01:00 |
|
|
|
a39d8300a1
|
Somewhere inbetween. Everything broken
|
2019-12-08 14:19:10 +01:00 |
|
|
|
3c0562cee4
|
Renamed nominals to nominals_txt
|
2018-09-21 09:16:58 +02:00 |
|
|
|
1b21cc1873
|
Partially done implementation of Octave and one-third octave filterbank.
|
2018-06-14 14:49:14 +02:00 |
|
|
|
ce7f31ca86
|
Renamed lasp_math.h to lasp_mat.h as it defines matrices, not math. Made a typedef from vd to dmat such that the types of these are the same. Same goes for vc and cmat
|
2018-04-01 14:01:07 +02:00 |
|
|
|
f37ce6b596
|
Fixed bug in filtebank, renamed ASCEE to LASP in lasp_python.h, added decimator.
|
2018-03-04 19:48:46 +01:00 |
|
|
|
53a60877b4
|
Cleanup fir filter design code. Removed Beamforming reference i test_fft.py
|
2018-02-27 19:45:21 +01:00 |
|
|
|
c55d3b1e17
|
Rename all beamforming to lasp. In all source code.
|
2018-02-23 20:40:45 +01:00 |
|
|
|
ad6c3213cd
|
Added filterbank and toebehoren
|
2018-02-19 10:40:28 +01:00 |
|
|
|
aa3935a82b
|
Fixed bug in print_dmat and print_cmat, allow for empty dmats and cmats in some part of the code. Added dmat_foreign_vd. Some minor improvements in code and docs. WARN and DBGWARN now print to stdout instead of stderr, as Spyder taps the stderr. Improved the dFifo to handle memory better. Now dFifo does not have a maximum size anymore. Renamed Fft_alloc to Fft_create.
|
2018-02-19 10:37:08 +01:00 |
|
|
|
008876dbab
|
Added inverse FFT code.
|
2018-02-17 17:01:53 +01:00 |
|
|
|
54e9eeca37
|
Changed internals of xmat to use strides instead of column pointers. Some other bugfixes
|
2018-02-12 20:44:07 +01:00 |
|
|
|
85636ce946
|
Changed Fft API to also be able to compute for one channel
|
2018-02-10 21:14:17 +01:00 |
|
|
|
89e9622b1f
|
Bugfix, added overflow checks for debug mode. Tested AvPowerSpectra. Now works without errors, only addTimeData cannot be called twice. Improved fft implementation. Removed double wrapper for fftpack. Renamed fft.pyx to wrappers.pyx. Improved CMakeLists. Adde proper function entrance and exit tracers. Made tracers indent to function level. Added Bartlett window function
|
2018-02-09 11:56:49 +01:00 |
|
|
|
83aaf69bef
|
Refactoring math routines. Added comments. Removed multithreading from fft routines. Switched to fftpack for FFT. Added Doxyfile
|
2018-02-06 12:01:27 +01:00 |
|
|
|
52e5a31bdb
|
Initial commit. Lots of stuff
|
2018-01-29 16:14:50 +01:00 |
|
