librosax.feature.tempogram_ratio¶

tempogram_ratio(*, y: ndarray | None = None, sr: float = 22050, onset_envelope: ndarray | None = None, tg: ndarray | None = None, bpm: ndarray | None = None, hop_length: int = 512, win_length: int = 384, start_bpm: float = 120, std_bpm: float = 1.0, max_tempo: float | None = 320.0, freqs: ndarray | None = None, factors: ndarray | None = None, aggregate: Callable | None = None, prior: Callable | None = None, center: bool = True, window: str = 'hann', kind: str = 'linear', fill_value: float = 0, norm: float | Callable | None = inf) → ndarray[source]¶

Compute tempogram ratio: the ratio of absolute energy in tempo harmonics.

This is a wrapper around librosa.feature.tempogram_ratio.

Parameters¶

ynp.ndarray [shape=(…, n)] or None: Audio time series (if tempogram not provided)
srfloat > 0: Sampling rate of the audio time series
onset_envelopenp.ndarray [shape=(n,)] or None: Pre-computed onset strength envelope (if tempogram not provided)
tgnp.ndarray [shape=(…, win_length, n)] or None: Pre-computed tempogram. If not provided, it will be computed from y or onset_envelope.
bpmnp.ndarray or None: Optional BPM values to evaluate
hop_lengthint > 0: Number of audio samples between successive onset measurements
win_lengthint > 0: Length of the tempogram window
start_bpmfloat > 0: Initial guess of the BPM
std_bpmfloat > 0: Standard deviation of tempo distribution
max_tempofloat > 0 or None: Maximum tempo to consider (in BPM)
freqsnp.ndarray or None: Optional frequency values
factorsnp.ndarray or None: Optional factors for ratio computation
aggregatecallable or None: Aggregation function
priorcallable or None: Prior distribution over tempo
centerbool: If True, tempogram windows are centered.
windowstr, tuple, number, callable, or list-like: A window specification
kindstr: Interpolation kind
fill_valuefloat: Fill value for interpolation
norm{np.inf, float > 0, None, callable}: Normalization mode