Spectrum is a Digital Spectral analysis software written in Python. The origin of this library is a C++ library that I wrote during my PhD, which was in C++ only. This former library was mainly based on the book "Digital Spectral analysis" from Marple Lawrence.
I re-engineered this work using Python so as to get rapidly a more robust code with a full set of documentation and tests.
For now, the library contains PSD estimates based on Fourier or Parametric methods. The Fourier methods are correlogram, periodogram and Welch estimates with a bunch of tapering windows from standard (Hann, Hamming, Blackman) to more exotic (DPSS, Taylor, ...). The parametric methods are based upon Yule-Walker and BURG estimates but also more complex such as coriance and modified covariance methods. Other methods based on eigen analysis such as MUSIC method are also implemented. In the close future, Multitapering methods should also be present