o sieB@sdZddlZddlZddlZddlmZddlZdZ dZ dZ dd Z d d Z dd dZddZddZd ddZddZddZddZddZdS)!a The goal of multiple f0 (multipitch) estimation and tracking is to identify all of the active fundamental frequencies in each time frame in a complex music signal. Conventions ----------- Multipitch estimates are represented by a timebase and a corresponding list of arrays of frequency estimates. Frequency estimates may have any number of frequency values, including 0 (represented by an empty array). Time values are in units of seconds and frequency estimates are in units of Hz. The timebase of the estimate time series should ideally match the timebase of the reference time series, but if this is not the case, the estimate time series is resampled using a nearest neighbor interpolation to match the estimate. Time values in the estimate time series that are outside of the range of the reference time series are given null (empty array) frequencies. By default, a frequency is "correct" if it is within 0.5 semitones of a reference frequency. Frequency values are compared by first mapping them to log-2 semitone space, where the distance between semitones is constant. Chroma-wrapped frequency values are computed by taking the log-2 frequency values modulo 12 to map them down to a single octave. A chroma-wrapped frequency estimate is correct if it's single-octave value is within 0.5 semitones of the single-octave reference frequency. The metrics are based on those described in [#poliner2007]_ and [#bay2009]_. Metrics ------- * :func:`mir_eval.multipitch.metrics`: Precision, Recall, Accuracy, Substitution, Miss, False Alarm, and Total Error scores based both on raw frequency values and values mapped to a single octave (chroma). References ---------- .. [#poliner2007] G. E. Poliner, and D. P. W. Ellis, "A Discriminative Model for Polyphonic Piano Transription", EURASIP Journal on Advances in Signal Processing, 2007(1):154-163, Jan. 2007. .. [#bay2009] Bay, M., Ehmann, A. F., & Downie, J. S. (2009). Evaluation of Multiple-F0 Estimation and Tracking Systems. In ISMIR (pp. 315-320). N)utilgL@g@g4@cCstj|tdtj|td|jdkrtd|jdkr!tdt|dkr,td|jdkr6td|jdkr?tdt|dkrJtd |jt|krUtd |jt|kr`td |D] }tj |t t d d qb|D] }tj |t t d d qpdS)apCheck that the time and frequency inputs are well-formed. Parameters ---------- ref_time : np.ndarray reference time stamps in seconds ref_freqs : list of np.ndarray reference frequencies in Hz est_time : np.ndarray estimate time stamps in seconds est_freqs : list of np.ndarray estimated frequencies in Hz )max_timerzReference times are empty.rz&Reference times have invalid dimensionz Reference frequencies are empty.zEstimated times are empty.z&Estimated times have invalid dimensionz Estimated frequencies are empty.z5Reference times and frequencies have unequal lengths.z4Estimate times and frequencies have unequal lengths.F)max_freqmin_freqallow_negativesN) rvalidate_eventsMAX_TIMEsizewarningswarnndim ValueErrorlenvalidate_frequenciesMAX_FREQMIN_FREQ)ref_time ref_freqsest_time est_freqsfreqrG/home/ubuntu/.local/lib/python3.10/site-packages/mir_eval/multipitch.pyvalidate9s6          rcs|jdkrgS|jdkrtggt|St|}td|}tjj||ddd|d|}|tggfdd|t D}|S)a(Resamples multipitch time series to a new timescale using nearest neighbor interpolation. Values in ``target_times`` outside the range of ``times`` return no pitch estimate. Parameters ---------- times : np.ndarray Array of time stamps frequencies : list of np.ndarray List of np.ndarrays of frequency values target_times : np.ndarray Array of target time stamps Returns ------- frequencies_resampled : list of numpy arrays Frequency list of lists resampled to new timebase rnearestFT)kind bounds_error assume_sorted fill_valuecsg|]}|qSrr).0i freq_valsrr sz'resample_multipitch..) r nparrayrarangescipy interpolateinterp1dastypeint)times frequencies target_timesn_timesfrequency_indexnew_frequency_indexfrequencies_resampledrr"rresample_multipitchfs&    r4{@csfdd|DS)aCConvert frequencies to continuous MIDI values. Parameters ---------- frequencies : list of np.ndarray Original frequency values ref_frequency : float reference frequency in Hz. Returns ------- frequencies_midi : list of np.ndarray Continuous MIDI frequency values. cs"g|] }ddt|qS)g@Q@g(@)r%log2r freqs ref_frequencyrrr$s"z'frequencies_to_midi..r)r.r:rr9rfrequencies_to_midisr;cCsdd|DS)a Wrap MIDI frequencies to a single octave (chroma). Parameters ---------- frequencies_midi : list of np.ndarray Continuous MIDI note frequency values. Returns ------- frequencies_chroma : list of np.ndarray Midi values wrapped to one octave. cSsg|]}t|dqS) )r%modr7rrrr$sz"midi_to_chroma..r)frequencies_midirrrmidi_to_chromasr?cCstdd|DS)zCompute the number of frequencies for each time point. Parameters ---------- frequencies : list of np.ndarray Frequency values Returns ------- num_freqs : np.ndarray Number of frequencies at each time point. cSsg|]}|jqSr)r )r frrrr$sz%compute_num_freqs..)r%r&)r.rrrcompute_num_freqss rA?Fc Csht|}t|f}tt||D] \}\}}|r$tj|||tjd} nt|||} t| ||<q|S)aCompute the number of true positives in an estimate given a reference. A frequency is correct if it is within a quartertone of the correct frequency. Parameters ---------- ref_freqs : list of np.ndarray reference frequencies (MIDI) est_freqs : list of np.ndarray estimated frequencies (MIDI) window : float Window size, in semitones chroma : bool If True, computes distances modulo n. If True, ``ref_freqs`` and ``est_freqs`` should be wrapped modulo n. Returns ------- true_positives : np.ndarray Array the same length as ref_freqs containing the number of true positives. )distance)rr%zeros enumeratezipr match_events_outer_distance_mod_n) rrwindowchroman_framestrue_positivesr! ref_frame est_framematchingrrrcompute_num_true_positivess  rPc Cst|}|}|dkr||}ntdd}|}|dkr+||}ntdd}|||}|dkrC||} nd} ||| fS)akCompute accuracy metrics. Parameters ---------- true_positives : np.ndarray Array containing the number of true positives at each time point. n_ref : np.ndarray Array containing the number of reference frequencies at each time point. n_est : np.ndarray Array containing the number of estimate frequencies at each time point. Returns ------- precision : float ``sum(true_positives)/sum(n_est)`` recall : float ``sum(true_positives)/sum(n_ref)`` acc : float ``sum(true_positives)/sum(n_est + n_ref - true_positives)`` rz#Estimate frequencies are all empty.$Reference frequencies are all empty.)floatsumr r ) rLn_refn_esttrue_positive_sum n_est_sum precision n_ref_sumrecall acc_denomaccrrrcompute_accuracys     r^c Cst|}|dkrtddStj||gdd||}||}d||dk<||}||}d||dk<||}tj||gdd||} |||| fS)a?Compute error score metrics. Parameters ---------- true_positives : np.ndarray Array containing the number of true positives at each time point. n_ref : np.ndarray Array containing the number of reference frequencies at each time point. n_est : np.ndarray Array containing the number of estimate frequencies at each time point. Returns ------- e_sub : float Substitution error e_miss : float Miss error e_fa : float False alarm error e_tot : float Total error rrR)rQrQrQrQ)axis)rSrTr r r%minmax) rLrUrVrZe_sube_miss_numeratore_misse_fa_numeratore_fae_totrrrcompute_err_score)s       rhcKst|||||j|jkst||stdt|||}t|}t|}t|}t|}t |} t |} t j t ||fi|} t j t ||fddi|} t | | | \} }}t| | | \}}}}t | | | \}}}t| | | \}}}}| |||||||||||||fS)aCompute multipitch metrics. All metrics are computed at the 'macro' level such that the frame true positive/false positive/false negative rates are summed across time and the metrics are computed on the combined values. Examples -------- >>> ref_time, ref_freqs = mir_eval.io.load_ragged_time_series( ... 'reference.txt') >>> est_time, est_freqs = mir_eval.io.load_ragged_time_series( ... 'estimated.txt') >>> metris_tuple = mir_eval.multipitch.metrics( ... ref_time, ref_freqs, est_time, est_freqs) Parameters ---------- ref_time : np.ndarray Time of each reference frequency value ref_freqs : list of np.ndarray List of np.ndarrays of reference frequency values est_time : np.ndarray Time of each estimated frequency value est_freqs : list of np.ndarray List of np.ndarrays of estimate frequency values **kwargs Additional keyword arguments which will be passed to the appropriate metric or preprocessing functions. Returns ------- precision : float Precision (TP/(TP + FP)) recall : float Recall (TP/(TP + FN)) accuracy : float Accuracy (TP/(TP + FP + FN)) e_sub : float Substitution error e_miss : float Miss error e_fa : float False alarm error e_tot : float Total error precision_chroma : float Chroma precision recall_chroma : float Chroma recall accuracy_chroma : float Chroma accuracy e_sub_chroma : float Chroma substitution error e_miss_chroma : float Chroma miss error e_fa_chroma : float Chroma false alarm error e_tot_chroma : float Chroma total error zLEstimate times not equal to reference times. Resampling to common time base.rJT)rr r%allcloser r r4r;r?rAr filter_kwargsrPr^rh)rrrrkwargsref_freqs_midiest_freqs_midiref_freqs_chromaest_freqs_chromarUrVrLtrue_positives_chromarYr[accuracyrbrdrfrgprecision_chroma recall_chromaaccuracy_chroma e_sub_chroma e_miss_chroma e_fa_chroma e_tot_chromarrrmetrics]sb<    rycKszt}tjt||||fi|\|d<|d<|d<|d<|d<|d<|d<|d<|d <|d <|d <|d <|d <|d<|S)aEvaluate two multipitch (multi-f0) transcriptions, where the first is treated as the reference (ground truth) and the second as the estimate to be evaluated (prediction). Examples -------- >>> ref_time, ref_freq = mir_eval.io.load_ragged_time_series('ref.txt') >>> est_time, est_freq = mir_eval.io.load_ragged_time_series('est.txt') >>> scores = mir_eval.multipitch.evaluate(ref_time, ref_freq, ... est_time, est_freq) Parameters ---------- ref_time : np.ndarray Time of each reference frequency value ref_freqs : list of np.ndarray List of np.ndarrays of reference frequency values est_time : np.ndarray Time of each estimated frequency value est_freqs : list of np.ndarray List of np.ndarrays of estimate frequency values **kwargs Additional keyword arguments which will be passed to the appropriate metric or preprocessing functions. Returns ------- scores : dict Dictionary of scores, where the key is the metric name (str) and the value is the (float) score achieved. PrecisionRecallAccuracyzSubstitution Errorz Miss ErrorzFalse Alarm Errorz Total ErrorzChroma Precisionz Chroma RecallzChroma AccuracyzChroma Substitution ErrorzChroma Miss ErrorzChroma False Alarm ErrorzChroma Total Error) collections OrderedDictrrjry)rrrrrkscoresrrrevaluates$!r)r5)rBF)__doc__numpyr%r}scipy.interpolater(rr r rrrr4r;r?rArPr^rhryrrrrrs(, - 6 *04