analyzeAudio 0.0.14__tar.gz → 0.0.16__tar.gz

{analyzeaudio-0.0.14 → analyzeaudio-0.0.16}/PKG-INFO

@@ -1,6 +1,6 @@
-Metadata-Version: 2.2
+Metadata-Version: 2.4
 Name: analyzeAudio
-Version: 0.0.14
+Version: 0.0.16
 Summary: Measure one or more aspects of one or more audio files.
 Author-email: Hunter Hogan <HunterHogan@pm.me>
 License: CC-BY-NC-4.0
@@ -15,7 +15,6 @@ Classifier: Intended Audience :: End Users/Desktop
 Classifier: Intended Audience :: Science/Research
 Classifier: Intended Audience :: Information Technology
 Classifier: Intended Audience :: Other Audience
-Classifier: License :: Free for non-commercial use
 Classifier: Natural Language :: English
 Classifier: Operating System :: OS Independent
 Classifier: Programming Language :: Python
@@ -36,7 +35,6 @@ License-File: LICENSE
 Requires-Dist: cachetools
 Requires-Dist: librosa
 Requires-Dist: numpy
-Requires-Dist: optype[numpy]
 Requires-Dist: standard-aifc; python_version >= "3.13"
 Requires-Dist: standard-sunau; python_version >= "3.13"
 Requires-Dist: torch
@@ -44,10 +42,11 @@ Requires-Dist: torchmetrics[audio]
 Requires-Dist: tqdm
 Requires-Dist: Z0Z_tools
 Provides-Extra: testing
+Requires-Dist: pytest; extra == "testing"
 Requires-Dist: pytest-cov; extra == "testing"
 Requires-Dist: pytest-xdist; extra == "testing"
-Requires-Dist: pytest; extra == "testing"
 Requires-Dist: pyupgrade; extra == "testing"
+Dynamic: license-file
 
 # analyzeAudio
 
@@ -139,12 +138,45 @@ print(audioAspects['Chromagram']['analyzerParameters'])
 'Signal entropy': float
 'Spectral Bandwidth': NDArray[float64] # shape(..., 1, frames)
 'Spectral Bandwidth mean': float
-'Spectral centroid': NDArray[float64] # shape(channels, frames)
-'Spectral centroid mean': float
 'Spectral Centroid': NDArray[float64] # shape(..., 1, frames)
 'Spectral Centroid mean': float
 'Spectral Contrast': NDArray[float64] # shape(..., 7, frames)
 'Spectral Contrast mean': float
+'Spectral Flatness': NDArray[float64] # shape(..., 1, frames)
+'Spectral Flatness mean': float
+'SRMR': NDArray[float64] # shape(...)
+'SRMR mean': float
+'Tempo': NDArray[float64] # shape(...)
+'Tempo mean': float
+'Tempogram': NDArray[float64] # shape(..., 384, samples)
+'Tempogram mean': float
+'Zero-crossing rate': NDArray[float64] # shape(..., 1, frames)
+'Zero-crossing rate mean': float
+'Zero-crossings rate': float
+```
+
+### I had to revert back to these
+
+```python
+'Spectral centroid': float
+'Spectral crest': float
+'Spectral decrease': float
+'Spectral entropy': float
+'Spectral flatness': float
+'Spectral flux': float
+'Spectral kurtosis': float
+'Spectral rolloff': float
+'Spectral skewness': float
+'Spectral slope': float
+'Spectral spread': float
+'Spectral variance': float
+```
+
+### Removed (temporarily, I hope)
+
+```python
+'Spectral centroid': NDArray[float64] # shape(channels, frames)
+'Spectral centroid mean': float
 'Spectral crest': NDArray[float64] # shape(channels, frames)
 'Spectral crest mean': float
 'Spectral decrease': NDArray[float64] # shape(channels, frames)
@@ -153,8 +185,6 @@ print(audioAspects['Chromagram']['analyzerParameters'])
 'Spectral entropy mean': float
 'Spectral flatness': NDArray[float64] # shape(channels, frames)
 'Spectral flatness mean': float
-'Spectral Flatness': NDArray[float64] # shape(..., 1, frames)
-'Spectral Flatness mean': float
 'Spectral flux': NDArray[float64] # shape(channels, frames)
 'Spectral flux mean': float
 'Spectral kurtosis': NDArray[float64] # shape(channels, frames)
@@ -169,15 +199,6 @@ print(audioAspects['Chromagram']['analyzerParameters'])
 'Spectral spread mean': float
 'Spectral variance': NDArray[float64] # shape(channels, frames)
 'Spectral variance mean': float
-'SRMR': NDArray[float64] # shape(...)
-'SRMR mean': float
-'Tempo': NDArray[float64] # shape(...)
-'Tempo mean': float
-'Tempogram': NDArray[float64] # shape(..., 384, samples)
-'Tempogram mean': float
-'Zero-crossing rate': NDArray[float64] # shape(..., 1, frames)
-'Zero-crossing rate mean': float
-'Zero-crossings rate': float
 ```
 
 ## Installation
@@ -191,4 +212,13 @@ pip install analyzeAudio
 [![Static Badge](https://img.shields.io/badge/2011_August-Homeless_since-blue?style=flat)](https://HunterThinks.com/support)
 [![YouTube Channel Subscribers](https://img.shields.io/youtube/channel/subscribers/UC3Gx7kz61009NbhpRtPP7tw)](https://www.youtube.com/@HunterHogan)
 
-[![CC-BY-NC-4.0](https://github.com/hunterhogan/analyzeAudio/blob/main/CC-BY-NC-4.0.png)](https://creativecommons.org/licenses/by-nc/4.0/)
+## How to code
+
+Coding One Step at a Time:
+
+0. WRITE CODE.
+1. Don't write stupid code that's hard to revise.
+2. Write good code.
+3. When revising, write better code.
+
+[![CC-BY-NC-4.0](https://github.com/hunterhogan/analyzeAudio/blob/main/CC-BY-NC-4.0.svg)](https://creativecommons.org/licenses/by-nc/4.0/)

{analyzeaudio-0.0.14 → analyzeaudio-0.0.16}/README.md

@@ -88,12 +88,45 @@ print(audioAspects['Chromagram']['analyzerParameters'])
 'Signal entropy': float
 'Spectral Bandwidth': NDArray[float64] # shape(..., 1, frames)
 'Spectral Bandwidth mean': float
-'Spectral centroid': NDArray[float64] # shape(channels, frames)
-'Spectral centroid mean': float
 'Spectral Centroid': NDArray[float64] # shape(..., 1, frames)
 'Spectral Centroid mean': float
 'Spectral Contrast': NDArray[float64] # shape(..., 7, frames)
 'Spectral Contrast mean': float
+'Spectral Flatness': NDArray[float64] # shape(..., 1, frames)
+'Spectral Flatness mean': float
+'SRMR': NDArray[float64] # shape(...)
+'SRMR mean': float
+'Tempo': NDArray[float64] # shape(...)
+'Tempo mean': float
+'Tempogram': NDArray[float64] # shape(..., 384, samples)
+'Tempogram mean': float
+'Zero-crossing rate': NDArray[float64] # shape(..., 1, frames)
+'Zero-crossing rate mean': float
+'Zero-crossings rate': float
+```
+
+### I had to revert back to these
+
+```python
+'Spectral centroid': float
+'Spectral crest': float
+'Spectral decrease': float
+'Spectral entropy': float
+'Spectral flatness': float
+'Spectral flux': float
+'Spectral kurtosis': float
+'Spectral rolloff': float
+'Spectral skewness': float
+'Spectral slope': float
+'Spectral spread': float
+'Spectral variance': float
+```
+
+### Removed (temporarily, I hope)
+
+```python
+'Spectral centroid': NDArray[float64] # shape(channels, frames)
+'Spectral centroid mean': float
 'Spectral crest': NDArray[float64] # shape(channels, frames)
 'Spectral crest mean': float
 'Spectral decrease': NDArray[float64] # shape(channels, frames)
@@ -102,8 +135,6 @@ print(audioAspects['Chromagram']['analyzerParameters'])
 'Spectral entropy mean': float
 'Spectral flatness': NDArray[float64] # shape(channels, frames)
 'Spectral flatness mean': float
-'Spectral Flatness': NDArray[float64] # shape(..., 1, frames)
-'Spectral Flatness mean': float
 'Spectral flux': NDArray[float64] # shape(channels, frames)
 'Spectral flux mean': float
 'Spectral kurtosis': NDArray[float64] # shape(channels, frames)
@@ -118,15 +149,6 @@ print(audioAspects['Chromagram']['analyzerParameters'])
 'Spectral spread mean': float
 'Spectral variance': NDArray[float64] # shape(channels, frames)
 'Spectral variance mean': float
-'SRMR': NDArray[float64] # shape(...)
-'SRMR mean': float
-'Tempo': NDArray[float64] # shape(...)
-'Tempo mean': float
-'Tempogram': NDArray[float64] # shape(..., 384, samples)
-'Tempogram mean': float
-'Zero-crossing rate': NDArray[float64] # shape(..., 1, frames)
-'Zero-crossing rate mean': float
-'Zero-crossings rate': float
 ```
 
 ## Installation
@@ -140,4 +162,13 @@ pip install analyzeAudio
 [![Static Badge](https://img.shields.io/badge/2011_August-Homeless_since-blue?style=flat)](https://HunterThinks.com/support)
 [![YouTube Channel Subscribers](https://img.shields.io/youtube/channel/subscribers/UC3Gx7kz61009NbhpRtPP7tw)](https://www.youtube.com/@HunterHogan)
 
-[![CC-BY-NC-4.0](https://github.com/hunterhogan/analyzeAudio/blob/main/CC-BY-NC-4.0.png)](https://creativecommons.org/licenses/by-nc/4.0/)
+## How to code
+
+Coding One Step at a Time:
+
+0. WRITE CODE.
+1. Don't write stupid code that's hard to revise.
+2. Write good code.
+3. When revising, write better code.
+
+[![CC-BY-NC-4.0](https://github.com/hunterhogan/analyzeAudio/blob/main/CC-BY-NC-4.0.svg)](https://creativecommons.org/licenses/by-nc/4.0/)

{analyzeaudio-0.0.14 → analyzeaudio-0.0.16}/analyzeAudio/analyzersUseFilename.py

@@ -1,4 +1,5 @@
-from .pythonator import pythonizeFFprobe
+"""Analyzers that use the filename of an audio file to analyze its audio data."""
+from analyzeAudio.pythonator import pythonizeFFprobe
 from analyzeAudio import registrationAudioAspect, cacheAudioAnalyzers
 from os import PathLike
 from statistics import mean
@@ -40,7 +41,7 @@ def getSI_SDRmean(pathFilenameAlpha: str | PathLike[Any], pathFilenameBeta: str
 	return SI_SDRmean
 
 @cachetools.cached(cache=cacheAudioAnalyzers)
-def ffprobeShotgunAndCache(pathFilename: str | PathLike[Any]) -> dict[str, float | numpy.ndarray]:
+def ffprobeShotgunAndCache(pathFilename: str | PathLike[Any]) -> dict[str, float]:
 	# for lavfi amovie/movie, the colons after driveLetter letters need to be escaped twice.
 	pFn = pathlib.PureWindowsPath(pathFilename)
 	lavfiPathFilename = pFn.drive.replace(":", "\\\\:")+pathlib.PureWindowsPath(pFn.root,pFn.relative_to(pFn.anchor)).as_posix()
@@ -63,14 +64,14 @@ def ffprobeShotgunAndCache(pathFilename: str | PathLike[Any]) -> dict[str, float
 	stdoutFFprobe, _DISCARDstderr = systemProcessFFprobe.communicate()
 	FFprobeStructured = pythonizeFFprobe(stdoutFFprobe.decode('utf-8'))[-1]
 
-	dictionaryAspectsAnalyzed: dict[str, float | numpy.ndarray] = {}
+	dictionaryAspectsAnalyzed: dict[str, float] = {}
 	if 'aspectralstats' in FFprobeStructured:
 		for keyName in FFprobeStructured['aspectralstats']:
 			# No matter how many channels, each keyName is `numpy.ndarray[tuple[int, int], numpy.dtype[numpy.float64]]`
 			# where `tuple[int, int]` is (channel, frame)
 			# NOTE (as of this writing) `registrar` can only understand the generic class `numpy.ndarray` and not more specific typing
-			dictionaryAspectsAnalyzed[keyName] = FFprobeStructured['aspectralstats'][keyName]
-			# dictionaryAspectsAnalyzed[keyName] = numpy.mean(FFprobeStructured['aspectralstats'][keyName]).astype(float)
+			# dictionaryAspectsAnalyzed[keyName] = FFprobeStructured['aspectralstats'][keyName]
+			dictionaryAspectsAnalyzed[keyName] = numpy.mean(FFprobeStructured['aspectralstats'][keyName]).astype(float)
 	if 'r128' in FFprobeStructured:
 		for keyName in FFprobeStructured['r128']:
 			dictionaryAspectsAnalyzed[keyName] = FFprobeStructured['r128'][keyName][-1]
@@ -133,55 +134,55 @@ def analyzeLUFShigh(pathFilename: str | PathLike[Any]) -> float | None:
 	return ffprobeShotgunAndCache(pathFilename).get('LRA.high')
 
 @registrationAudioAspect('Power spectral density')
-def analyzeMean(pathFilename: str | PathLike[Any]) -> numpy.ndarray:
+def analyzeMean(pathFilename: str | PathLike[Any]) -> float | None:
 	return ffprobeShotgunAndCache(pathFilename).get('mean')
 
 @registrationAudioAspect('Spectral variance')
-def analyzeVariance(pathFilename: str | PathLike[Any]) -> numpy.ndarray:
+def analyzeVariance(pathFilename: str | PathLike[Any]) -> float | None:
 	return ffprobeShotgunAndCache(pathFilename).get('variance')
 
 @registrationAudioAspect('Spectral centroid')
-def analyzeCentroid(pathFilename: str | PathLike[Any]) -> numpy.ndarray:
+def analyzeCentroid(pathFilename: str | PathLike[Any]) -> float | None:
 	return ffprobeShotgunAndCache(pathFilename).get('centroid')
 
 @registrationAudioAspect('Spectral spread')
-def analyzeSpread(pathFilename: str | PathLike[Any]) -> numpy.ndarray:
+def analyzeSpread(pathFilename: str | PathLike[Any]) -> float | None:
 	return ffprobeShotgunAndCache(pathFilename).get('spread')
 
 @registrationAudioAspect('Spectral skewness')
-def analyzeSkewness(pathFilename: str | PathLike[Any]) -> numpy.ndarray:
+def analyzeSkewness(pathFilename: str | PathLike[Any]) -> float | None:
 	return ffprobeShotgunAndCache(pathFilename).get('skewness')
 
 @registrationAudioAspect('Spectral kurtosis')
-def analyzeKurtosis(pathFilename: str | PathLike[Any]) -> numpy.ndarray:
+def analyzeKurtosis(pathFilename: str | PathLike[Any]) -> float | None:
 	return ffprobeShotgunAndCache(pathFilename).get('kurtosis')
 
 @registrationAudioAspect('Spectral entropy')
-def analyzeSpectralEntropy(pathFilename: str | PathLike[Any]) -> numpy.ndarray:
+def analyzeSpectralEntropy(pathFilename: str | PathLike[Any]) -> float | None:
 	return ffprobeShotgunAndCache(pathFilename).get('entropy')
 
 @registrationAudioAspect('Spectral flatness')
-def analyzeFlatness(pathFilename: str | PathLike[Any]) -> numpy.ndarray:
+def analyzeFlatness(pathFilename: str | PathLike[Any]) -> float | None:
 	return ffprobeShotgunAndCache(pathFilename).get('flatness')
 
 @registrationAudioAspect('Spectral crest')
-def analyzeCrest(pathFilename: str | PathLike[Any]) -> numpy.ndarray:
+def analyzeCrest(pathFilename: str | PathLike[Any]) -> float | None:
 	return ffprobeShotgunAndCache(pathFilename).get('crest')
 
 @registrationAudioAspect('Spectral flux')
-def analyzeFlux(pathFilename: str | PathLike[Any]) -> numpy.ndarray:
+def analyzeFlux(pathFilename: str | PathLike[Any]) -> float | None:
 	return ffprobeShotgunAndCache(pathFilename).get('flux')
 
 @registrationAudioAspect('Spectral slope')
-def analyzeSlope(pathFilename: str | PathLike[Any]) -> numpy.ndarray:
+def analyzeSlope(pathFilename: str | PathLike[Any]) -> float | None:
 	return ffprobeShotgunAndCache(pathFilename).get('slope')
 
 @registrationAudioAspect('Spectral decrease')
-def analyzeDecrease(pathFilename: str | PathLike[Any]) -> numpy.ndarray:
+def analyzeDecrease(pathFilename: str | PathLike[Any]) -> float | None:
 	return ffprobeShotgunAndCache(pathFilename).get('decrease')
 
 @registrationAudioAspect('Spectral rolloff')
-def analyzeRolloff(pathFilename: str | PathLike[Any]) -> numpy.ndarray:
+def analyzeRolloff(pathFilename: str | PathLike[Any]) -> float | None:
 	return ffprobeShotgunAndCache(pathFilename).get('rolloff')
 
 @registrationAudioAspect('Abs_Peak_count')

{analyzeaudio-0.0.14 → analyzeaudio-0.0.16}/analyzeAudio/analyzersUseSpectrogram.py

@@ -1,30 +1,30 @@
+"""Analyzers that use the spectrogram to analyze audio data."""
 from analyzeAudio import registrationAudioAspect, audioAspects, cacheAudioAnalyzers
 from typing import Any
 import cachetools
 import librosa
 import numpy
-from optype.numpy import AnyFloatingDType, ToArray3D, ToFloat3D
 from numpy import dtype, floating
 
 @registrationAudioAspect('Chromagram')
-def analyzeChromagram(spectrogramPower: numpy.ndarray[Any, dtype[floating[Any]]], sampleRate: int, **keywordArguments: Any) -> numpy.ndarray:
-	return librosa.feature.chroma_stft(S=spectrogramPower, sr=sampleRate, **keywordArguments)
+def analyzeChromagram(spectrogramPower: numpy.ndarray[Any, dtype[floating[Any]]], sampleRate: int, **keywordArguments: Any) -> numpy.ndarray: # pyright: ignore [reportMissingTypeArgument, reportUnknownParameterType]
+	return librosa.feature.chroma_stft(S=spectrogramPower, sr=sampleRate, **keywordArguments) # type: ignore
 
 @registrationAudioAspect('Spectral Contrast')
-def analyzeSpectralContrast(spectrogramMagnitude: numpy.ndarray[Any, dtype[floating[Any]]], **keywordArguments: Any) -> numpy.ndarray:
-	return librosa.feature.spectral_contrast(S=spectrogramMagnitude, **keywordArguments)
+def analyzeSpectralContrast(spectrogramMagnitude: numpy.ndarray[Any, dtype[floating[Any]]], **keywordArguments: Any) -> numpy.ndarray: # pyright: ignore [reportMissingTypeArgument, reportUnknownParameterType]
+	return librosa.feature.spectral_contrast(S=spectrogramMagnitude, **keywordArguments) # type: ignore
 
 @registrationAudioAspect('Spectral Bandwidth')
-def analyzeSpectralBandwidth(spectrogramMagnitude: numpy.ndarray[Any, dtype[floating[Any]]], **keywordArguments: Any) -> numpy.ndarray:
+def analyzeSpectralBandwidth(spectrogramMagnitude: numpy.ndarray[Any, dtype[floating[Any]]], **keywordArguments: Any) -> numpy.ndarray: # pyright: ignore [reportMissingTypeArgument, reportUnknownParameterType]
 	centroid = audioAspects['Spectral Centroid']['analyzer'](spectrogramMagnitude)
-	return librosa.feature.spectral_bandwidth(S=spectrogramMagnitude, centroid=centroid, **keywordArguments)
+	return librosa.feature.spectral_bandwidth(S=spectrogramMagnitude, centroid=centroid, **keywordArguments) # type: ignore
 
 @cachetools.cached(cache=cacheAudioAnalyzers)
 @registrationAudioAspect('Spectral Centroid')
-def analyzeSpectralCentroid(spectrogramMagnitude: numpy.ndarray[Any, dtype[floating[Any]]], **keywordArguments: Any) -> numpy.ndarray:
-	return librosa.feature.spectral_centroid(S=spectrogramMagnitude, **keywordArguments)
+def analyzeSpectralCentroid(spectrogramMagnitude: numpy.ndarray[Any, dtype[floating[Any]]], **keywordArguments: Any) -> numpy.ndarray: # pyright: ignore [reportMissingTypeArgument, reportUnknownParameterType]
+	return librosa.feature.spectral_centroid(S=spectrogramMagnitude, **keywordArguments) # type: ignore
 
 @registrationAudioAspect('Spectral Flatness')
-def analyzeSpectralFlatness(spectrogramMagnitude: numpy.ndarray[Any, dtype[floating[Any]]], **keywordArguments: Any) -> numpy.ndarray:
-	spectralFlatness = librosa.feature.spectral_flatness(S=spectrogramMagnitude, **keywordArguments)
+def analyzeSpectralFlatness(spectrogramMagnitude: numpy.ndarray[Any, dtype[floating[Any]]], **keywordArguments: Any) -> numpy.ndarray: # pyright: ignore [reportMissingTypeArgument, reportUnknownParameterType]
+	spectralFlatness: numpy.ndarray[tuple[int, ...], numpy.dtype[numpy.floating[Any]]] = librosa.feature.spectral_flatness(S=spectrogramMagnitude, **keywordArguments) # type: ignore
 	return 20 * numpy.log10(spectralFlatness, where=(spectralFlatness != 0)) # dB

{analyzeaudio-0.0.14 → analyzeaudio-0.0.16}/analyzeAudio/analyzersUseTensor.py

@@ -1,3 +1,4 @@
+"""Analyzers that use the tensor to analyze audio data."""
 from analyzeAudio import registrationAudioAspect
 from torchmetrics.functional.audio.srmr import speech_reverberation_modulation_energy_ratio
 from typing import Any
@@ -5,6 +6,6 @@ import numpy
 import torch
 
 @registrationAudioAspect('SRMR')
-def analyzeSRMR(tensorAudio: torch.Tensor, sampleRate: int, pytorchOnCPU: bool | None, **keywordArguments: Any) -> numpy.ndarray:
+def analyzeSRMR(tensorAudio: torch.Tensor, sampleRate: int, pytorchOnCPU: bool | None, **keywordArguments: Any) -> numpy.ndarray: # pyright: ignore [reportMissingTypeArgument, reportUnknownParameterType]
 	keywordArguments['fast'] = keywordArguments.get('fast') or pytorchOnCPU or None
-	return torch.Tensor.numpy(speech_reverberation_modulation_energy_ratio(tensorAudio, sampleRate, **keywordArguments))
+	return torch.Tensor.numpy(speech_reverberation_modulation_energy_ratio(tensorAudio, sampleRate, **keywordArguments)) # type: ignore

analyzeaudio-0.0.16/analyzeAudio/analyzersUseWaveform.py

@@ -0,0 +1,26 @@
+"""Analyzers that use the waveform of audio data."""
+from analyzeAudio import registrationAudioAspect, audioAspects, cacheAudioAnalyzers
+from typing import Any
+import librosa
+import numpy
+import cachetools
+
+@cachetools.cached(cache=cacheAudioAnalyzers)
+@registrationAudioAspect('Tempogram')
+def analyzeTempogram(waveform: numpy.ndarray[tuple[int, ...], numpy.dtype[numpy.floating[Any]]], sampleRate: int, **keywordArguments: Any) -> numpy.ndarray: # pyright: ignore [reportMissingTypeArgument, reportUnknownParameterType]
+	return librosa.feature.tempogram(y=waveform, sr=sampleRate, **keywordArguments) # type: ignore
+
+# "RMS value from audio samples is faster ... However, ... spectrogram ... more accurate ... because ... windowed"
+@registrationAudioAspect('RMS from waveform')
+def analyzeRMS(waveform: numpy.ndarray[tuple[int, ...], numpy.dtype[numpy.floating[Any]]], **keywordArguments: Any) -> numpy.ndarray: # pyright: ignore [reportMissingTypeArgument, reportUnknownParameterType]
+	arrayRMS: numpy.ndarray[tuple[int, ...], numpy.dtype[numpy.floating[Any]]] = librosa.feature.rms(y=waveform, **keywordArguments) # type: ignore
+	return 20 * numpy.log10(arrayRMS, where=(arrayRMS != 0)) # dB
+
+@registrationAudioAspect('Tempo')
+def analyzeTempo(waveform: numpy.ndarray[tuple[int, ...], numpy.dtype[numpy.floating[Any]]], sampleRate: int, **keywordArguments: Any) -> numpy.ndarray: # pyright: ignore [reportMissingTypeArgument, reportUnknownParameterType]
+	tempogram = audioAspects['Tempogram']['analyzer'](waveform, sampleRate)
+	return librosa.feature.tempo(y=waveform, sr=sampleRate, tg=tempogram, **keywordArguments) # type: ignore
+
+@registrationAudioAspect('Zero-crossing rate') # This is distinct from 'Zero-crossings rate'
+def analyzeZeroCrossingRate(waveform: numpy.ndarray[tuple[int, ...], numpy.dtype[numpy.floating[Any]]], **keywordArguments: Any) -> numpy.ndarray: # pyright: ignore [reportMissingTypeArgument, reportUnknownParameterType]
+	return librosa.feature.zero_crossing_rate(y=waveform, **keywordArguments) # type: ignore

{analyzeaudio-0.0.14 → analyzeaudio-0.0.16}/analyzeAudio/audioAspectsRegistry.py

@@ -1,14 +1,12 @@
 from collections.abc import Callable, Sequence
 from concurrent.futures import ProcessPoolExecutor, as_completed
 from numpy.typing import NDArray
+from os import PathLike
 from typing import Any, cast, ParamSpec, TypeAlias, TYPE_CHECKING, TypeVar
-from Z0Z_tools import defineConcurrencyLimit, oopsieKwargsie, stft
+from Z0Z_tools import defineConcurrencyLimit, oopsieKwargsie, stft, Spectrogram
 import cachetools
 import inspect
-import librosa
-import multiprocessing
 import numpy
-from os import PathLike
 import pathlib
 import soundfile
 import torch
@@ -19,8 +17,11 @@ if TYPE_CHECKING:
 else:
 	TypedDict = dict
 
-if __name__ == '__main__':
-	multiprocessing.set_start_method('spawn')
+from multiprocessing import set_start_method as multiprocessing_set_start_method
+try:
+	multiprocessing_set_start_method('spawn')
+except RuntimeError:
+	pass
 
 warnings.filterwarnings('ignore', category=UserWarning, module='torchmetrics', message='.*fast=True.*')
 
@@ -101,7 +102,7 @@ def analyzeAudioFile(pathFilename: str | PathLike[Any], listAspectNames: list[st
 
 	# I need "lazy" loading
 	tryAgain = True
-	while tryAgain: # `tenacity`?
+	while tryAgain:
 		try:
 			tensorAudio = torch.from_numpy(waveform)  # memory-sharing
 			tryAgain = False

{analyzeaudio-0.0.14 → analyzeaudio-0.0.16}/analyzeAudio/pythonator.py

@@ -1,3 +1,4 @@
+"""Convert FFprobe output to a standardized Python object."""
 from collections import defaultdict
 from typing import Any, cast, NamedTuple
 import json
@@ -89,7 +90,7 @@ def pythonizeFFprobe(FFprobeJSON_utf8: str):
 											Z0Z_dictionaries[registrant] = {}
 										elif statistic not in Z0Z_dictionaries[registrant]:
 												# NOTE (as of this writing) `registrar` can only understand the generic class `numpy.ndarray` and not more specific typing
-												valueSherpa = cast(numpy.ndarray, numpy.zeros((channel, len(FFroot['frames']))))
+												valueSherpa = cast(numpy.ndarray, numpy.zeros((channel, len(FFroot['frames'])))) # type: ignore
 												Z0Z_dictionaries[registrant][statistic] = valueSherpa
 										else:
 											raise  # Re-raise the exception

{analyzeaudio-0.0.14 → analyzeaudio-0.0.16}/analyzeAudio.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
-Metadata-Version: 2.2
+Metadata-Version: 2.4
 Name: analyzeAudio
-Version: 0.0.14
+Version: 0.0.16
 Summary: Measure one or more aspects of one or more audio files.
 Author-email: Hunter Hogan <HunterHogan@pm.me>
 License: CC-BY-NC-4.0
@@ -15,7 +15,6 @@ Classifier: Intended Audience :: End Users/Desktop
 Classifier: Intended Audience :: Science/Research
 Classifier: Intended Audience :: Information Technology
 Classifier: Intended Audience :: Other Audience
-Classifier: License :: Free for non-commercial use
 Classifier: Natural Language :: English
 Classifier: Operating System :: OS Independent
 Classifier: Programming Language :: Python
@@ -36,7 +35,6 @@ License-File: LICENSE
 Requires-Dist: cachetools
 Requires-Dist: librosa
 Requires-Dist: numpy
-Requires-Dist: optype[numpy]
 Requires-Dist: standard-aifc; python_version >= "3.13"
 Requires-Dist: standard-sunau; python_version >= "3.13"
 Requires-Dist: torch
@@ -44,10 +42,11 @@ Requires-Dist: torchmetrics[audio]
 Requires-Dist: tqdm
 Requires-Dist: Z0Z_tools
 Provides-Extra: testing
+Requires-Dist: pytest; extra == "testing"
 Requires-Dist: pytest-cov; extra == "testing"
 Requires-Dist: pytest-xdist; extra == "testing"
-Requires-Dist: pytest; extra == "testing"
 Requires-Dist: pyupgrade; extra == "testing"
+Dynamic: license-file
 
 # analyzeAudio
 
@@ -139,12 +138,45 @@ print(audioAspects['Chromagram']['analyzerParameters'])
 'Signal entropy': float
 'Spectral Bandwidth': NDArray[float64] # shape(..., 1, frames)
 'Spectral Bandwidth mean': float
-'Spectral centroid': NDArray[float64] # shape(channels, frames)
-'Spectral centroid mean': float
 'Spectral Centroid': NDArray[float64] # shape(..., 1, frames)
 'Spectral Centroid mean': float
 'Spectral Contrast': NDArray[float64] # shape(..., 7, frames)
 'Spectral Contrast mean': float
+'Spectral Flatness': NDArray[float64] # shape(..., 1, frames)
+'Spectral Flatness mean': float
+'SRMR': NDArray[float64] # shape(...)
+'SRMR mean': float
+'Tempo': NDArray[float64] # shape(...)
+'Tempo mean': float
+'Tempogram': NDArray[float64] # shape(..., 384, samples)
+'Tempogram mean': float
+'Zero-crossing rate': NDArray[float64] # shape(..., 1, frames)
+'Zero-crossing rate mean': float
+'Zero-crossings rate': float
+```
+
+### I had to revert back to these
+
+```python
+'Spectral centroid': float
+'Spectral crest': float
+'Spectral decrease': float
+'Spectral entropy': float
+'Spectral flatness': float
+'Spectral flux': float
+'Spectral kurtosis': float
+'Spectral rolloff': float
+'Spectral skewness': float
+'Spectral slope': float
+'Spectral spread': float
+'Spectral variance': float
+```
+
+### Removed (temporarily, I hope)
+
+```python
+'Spectral centroid': NDArray[float64] # shape(channels, frames)
+'Spectral centroid mean': float
 'Spectral crest': NDArray[float64] # shape(channels, frames)
 'Spectral crest mean': float
 'Spectral decrease': NDArray[float64] # shape(channels, frames)
@@ -153,8 +185,6 @@ print(audioAspects['Chromagram']['analyzerParameters'])
 'Spectral entropy mean': float
 'Spectral flatness': NDArray[float64] # shape(channels, frames)
 'Spectral flatness mean': float
-'Spectral Flatness': NDArray[float64] # shape(..., 1, frames)
-'Spectral Flatness mean': float
 'Spectral flux': NDArray[float64] # shape(channels, frames)
 'Spectral flux mean': float
 'Spectral kurtosis': NDArray[float64] # shape(channels, frames)
@@ -169,15 +199,6 @@ print(audioAspects['Chromagram']['analyzerParameters'])
 'Spectral spread mean': float
 'Spectral variance': NDArray[float64] # shape(channels, frames)
 'Spectral variance mean': float
-'SRMR': NDArray[float64] # shape(...)
-'SRMR mean': float
-'Tempo': NDArray[float64] # shape(...)
-'Tempo mean': float
-'Tempogram': NDArray[float64] # shape(..., 384, samples)
-'Tempogram mean': float
-'Zero-crossing rate': NDArray[float64] # shape(..., 1, frames)
-'Zero-crossing rate mean': float
-'Zero-crossings rate': float
 ```
 
 ## Installation
@@ -191,4 +212,13 @@ pip install analyzeAudio
 [![Static Badge](https://img.shields.io/badge/2011_August-Homeless_since-blue?style=flat)](https://HunterThinks.com/support)
 [![YouTube Channel Subscribers](https://img.shields.io/youtube/channel/subscribers/UC3Gx7kz61009NbhpRtPP7tw)](https://www.youtube.com/@HunterHogan)
 
-[![CC-BY-NC-4.0](https://github.com/hunterhogan/analyzeAudio/blob/main/CC-BY-NC-4.0.png)](https://creativecommons.org/licenses/by-nc/4.0/)
+## How to code
+
+Coding One Step at a Time:
+
+0. WRITE CODE.
+1. Don't write stupid code that's hard to revise.
+2. Write good code.
+3. When revising, write better code.
+
+[![CC-BY-NC-4.0](https://github.com/hunterhogan/analyzeAudio/blob/main/CC-BY-NC-4.0.svg)](https://creativecommons.org/licenses/by-nc/4.0/)

{analyzeaudio-0.0.14 → analyzeaudio-0.0.16}/analyzeAudio.egg-info/SOURCES.txt

@@ -7,6 +7,7 @@ analyzeAudio/analyzersUseSpectrogram.py
 analyzeAudio/analyzersUseTensor.py
 analyzeAudio/analyzersUseWaveform.py
 analyzeAudio/audioAspectsRegistry.py
+analyzeAudio/py.typed
 analyzeAudio/pythonator.py
 analyzeAudio.egg-info/PKG-INFO
 analyzeAudio.egg-info/SOURCES.txt

{analyzeaudio-0.0.14 → analyzeaudio-0.0.16}/analyzeAudio.egg-info/requires.txt

@@ -1,7 +1,6 @@
 cachetools
 librosa
 numpy
-optype[numpy]
 torch
 torchmetrics[audio]
 tqdm
@@ -12,7 +11,7 @@ standard-aifc
 standard-sunau
 
 [testing]
+pytest
 pytest-cov
 pytest-xdist
-pytest
 pyupgrade

{analyzeaudio-0.0.14 → analyzeaudio-0.0.16}/pyproject.toml

@@ -12,7 +12,6 @@ classifiers = [
     "Intended Audience :: Science/Research",
     "Intended Audience :: Information Technology",
     "Intended Audience :: Other Audience",
-    "License :: Free for non-commercial use",
     "Natural Language :: English",
     "Operating System :: OS Independent",
     "Programming Language :: Python",
@@ -31,7 +30,6 @@ dependencies = [
     "cachetools",
     "librosa",
     "numpy",
-    "optype[numpy]",
     "standard-aifc;python_version>='3.13'",
     "standard-sunau;python_version>='3.13'",
     "torch",
@@ -43,14 +41,14 @@ keywords = ["audio", "analysis", "measurement", "metrics", "torch", "spectrum",
 license = { 'text' = "CC-BY-NC-4.0" }
 name = "analyzeAudio"
 optional-dependencies = { testing = [
+    "pytest",
     "pytest-cov",
     "pytest-xdist",
-    "pytest",
     "pyupgrade",] }
 readme = { file = "README.md", content-type = "text/markdown" }
 requires-python = ">=3.10"
 urls = { Donate = "https://www.patreon.com/integrated", Homepage = "https://github.com/hunterhogan/analyzeAudio", Repository = "https://github.com/hunterhogan/analyzeAudio.git"  }
-version = "0.0.14"
+version = "0.0.16"
 
 [project.scripts]
 whatMeasurements = "analyzeAudio.audioAspectsRegistry:getListAvailableAudioAspects"
@@ -69,12 +67,13 @@ run = { branch = true, concurrency = [
 xml = { output = "tests/coverage/coverage.xml" }
 
 [tool.pytest.ini_options]
-log_auto_indent = ["On"]
-addopts = ["--color=yes", "-n 4"]
+log_auto_indent = true
+addopts = ["--color=auto", "-n 4"]
 testpaths = ["tests"]
 
-[tool.setuptools.packages.find]
-where = ["."]
+[tool.setuptools]
+package-data = { "*" = ["py.typed"] }
+packages = { find = {} }
 
 [tool.updateCitation]
 filenameCitationDOTcff = 'CITATION.cff'

analyzeaudio-0.0.14/analyzeAudio/analyzersUseWaveform.py

@@ -1,26 +0,0 @@
-from analyzeAudio import registrationAudioAspect, audioAspects, cacheAudioAnalyzers
-from typing import Any
-import librosa
-import numpy
-from optype.numpy import ToArray2D, AnyFloatingDType
-import cachetools
-
-@cachetools.cached(cache=cacheAudioAnalyzers)
-@registrationAudioAspect('Tempogram')
-def analyzeTempogram(waveform: ToArray2D[AnyFloatingDType], sampleRate: int, **keywordArguments: Any) -> numpy.ndarray:
-	return librosa.feature.tempogram(y=waveform, sr=sampleRate, **keywordArguments)
-
-# "RMS value from audio samples is faster ... However, ... spectrogram ... more accurate ... because ... windowed"
-@registrationAudioAspect('RMS from waveform')
-def analyzeRMS(waveform: ToArray2D[AnyFloatingDType], **keywordArguments: Any) -> numpy.ndarray:
-	arrayRMS = librosa.feature.rms(y=waveform, **keywordArguments)
-	return 20 * numpy.log10(arrayRMS, where=(arrayRMS != 0)) # dB
-
-@registrationAudioAspect('Tempo')
-def analyzeTempo(waveform: ToArray2D[AnyFloatingDType], sampleRate: int, **keywordArguments: Any) -> numpy.ndarray:
-	tempogram = audioAspects['Tempogram']['analyzer'](waveform, sampleRate)
-	return librosa.feature.tempo(y=waveform, sr=sampleRate, tg=tempogram, **keywordArguments)
-
-@registrationAudioAspect('Zero-crossing rate') # This is distinct from 'Zero-crossings rate'
-def analyzeZeroCrossingRate(waveform: ToArray2D[AnyFloatingDType], **keywordArguments: Any) -> numpy.ndarray:
-	return librosa.feature.zero_crossing_rate(y=waveform, **keywordArguments)