RFA v1.0

RFA implements the RFA (Rhythm Formant Analysis) method for analysing rhythms and rhythm variation in speech using spectrum and spectrogram analysis. The method is used for comparing recordings and displaying difference/similarity relations in distance maps based on standard distance metrics and as hierarchical clustering in similarity dendrograms.
The implementation is in Python 3.n using a small set of standard Python libraries. Three applications are included, with documentation in the code of each application:

1. M_AM_FM_LF_spectrogram_formants.py for single speech files, for displaying and outputting CSV files and PNG graphics of amplitude envelope modulation, f0 modulation, and the spectra and spectrograms associated with these two modulation types.
2. N_AM_FM_LF_spectrum_distance_networks.py for batches of speech files, with pairwise comparison of files based on spectrum vectors and vectors of peak values in the individual spectra of a spectrogram.
   Custom local modules are provided for access by these applications:
3. module_distancenetworks.py
4. module_drawdendrogram.py
5. module_fm_demodulation.py
6. module_hierarchicalclustering.py
7. module_spectrogram.py
   Development and usage notes are included in each of the application files.
   If RFA is used in publications, the following publication must be cited:
   Gibbon, D. (2021). The rhythms of rhythm. Journal of the International Phonetic Association, 1-33. doi:10.1017/S0025100321000086
   GPL 4.0