mvronkin/dsatools

Digital signal analysis library for python. The library includes such methods of the signal analysis, signal processing and signal parameter estimation as ARMA-based techniques; subspace-based techniques; matrix-pencil-based methods; singular-spectrum analysis (SSA); dynamic-mode decomposition (DMD); empirical mode decomposition; variational mode decomposition (EMD); empirical wavelet transform (EWT); Hilbert vibration decomposition (HVD) and many others.

Digital Signal Analysis (DSA) library for python. The library includes such methods of signal analysis and signal parameter estimation as arma-based techniques; subspace-based techniques; matrix-pencil-based methods; singular-spectrum analysis (SSA); dynamic-mode decomposition (DMD); empirical mode decomposition; variational mode decomposition; empirical wavelet transform; Hilbert vibration decomposition and many others.

For install use

pip install dsatools

For cite you may use

Ronkin, Mikhail V., et al. "Numerical analysis of adaptive signal decomposition methods applied for ultrasonic gas flowmeters." AIP Conference Proceedings. Vol. 2425. No. 1. AIP Publishing LLC, 2022.

Google schoar; research gate

bibtex

@inproceedings{ronkin2022numerical,
 title={Numerical analysis of adaptive signal decomposition methods applied for ultrasonic gas flowmeters},
 author={Ronkin, Mikhail V and Kalmykov, Alexey A and Polyakov, Stanislav O and Nagovicin, Viktor S},
 booktitle={AIP Conference Proceedings},
 volume={2425},
 number={1},
 pages={130009},
 year={2022},
 organization={AIP Publishing LLC}
}

Also paper on the topic of using dsatools: Ronkin, Mikhail, and Dima Bykhovsky. "Passive Fingerprinting of Same-Model Electrical Devices by Current Consumption." Sensors 23, no. 1 (2023): 533.

Google schoar; research gate; mdpi

bibtex

@article{ronkin2023passive,
 title={Passive Fingerprinting of Same-Model Electrical Devices by Current Consumption},
 author={Ronkin, Mikhail and Bykhovsky, Dima},
 journal={Sensors},
 volume={23},
 number={1},
 pages={533},
 year={2023},
 publisher={MDPI}
}

dsatools's Issues

Decomposition

Why when applying the decomposition function, let's say the EMD func, the sum result is not equal to the original signal?

setup.py install fails due to missing README.md file

Hi!
Thanks for creating such a useful library.
One small issue I found is that setup.py install fails in my environment.

$ python setup.py install                 
Traceback (most recent call last):
  File "/home/diarmaid/Documents/Dev/driverDrowsiness/other/dsatools/setup.py", line 4, in <module>
    with open('README.md') as f:
FileNotFoundError: [Errno 2] No such file or directory: 'README.md'

Creating a temporary README.md file solves the issue, the install can complete.

Regararding LPF value for smoothing

Thank you for your work on HVD.

I am having a question regarding the Low Pass Filtering value which acts as a smoothing factor for HVD decomposition. In one of the references you mentioned, the LPF value is usually very low around 0.02 to 0.05Hz. However, I see your implementation here is different with values up to 20 and 30. What does these values mean?

Missing utilities import in ./operators/_hht.py file

Just a minor issue, there appears to be a missing import in the ./operators/_hht.py file.
`---------------------------------------------------------------------------
NameError Traceback (most recent call last)
/tmp/ipykernel_406881/1073320869.py in
29 psds = decomposition.hvd(window_analytical, order=1, fpar=30, ret_freqs=False)
30 # print(psds * 500)
---> 31 instantfreqs = dsatools.operators.hht(psds)
32 style.use('seaborn-dark') # So that we can see the labels on a dark bg
33 fig, ax = plt.subplots(2, 1, figsize=(25,10))

~/anaconda3/envs/auto_labeller/lib/python3.9/site-packages/dsatools-0.1.44-py3.9.egg/dsatools/operators/_hht.py in hht(components)
57
58 #TODO: optimize, uzing explicity unwrap rutine
---> 59 instantfreqs[i,:] = ut.diff(np.unwrap(np.angle(component)))/2/np.pi
60 return envelopes, instantfreqs

NameError: name 'ut' is not defined`

Adding the import:

import dsatools.utilits as ut
Fixes this.