Feature Extractor

You can use FeatureExtractor for batch feature extraction

# Get a 880Hz's audio file
import audioflux as af
from audioflux.type import SpectralFilterBankScaleType
sample_path = af.utils.sample_path('880')
audio_arr, sr = af.read(sample_path)

# Create FeatureExtractor object and extract spectrogram
fa_obj = af.FeatureExtractor(transforms=['bft', 'cwt', 'cqt'], samplate=sr, radix2_exp=12,
                             scale_type=SpectralFilterBankScaleType.OCTAVE)
spec_result = fa_obj.spectrogram(audio_arr, is_continue=True)

#Extract spectral/xxcc/deconv
spectral_result = fa_obj.spectral(spec_result, spectral='flux',
                                  spectral_kw={'is_positive': True})
xxcc_result = fa_obj.xxcc(spec_result, cc_num=13)
deconv_result = fa_obj.deconv(spec_result)