Index: librosa-0.9.2/librosa/segment.py =================================================================== --- librosa-0.9.2.orig/librosa/segment.py +++ librosa-0.9.2/librosa/segment.py @@ -255,7 +255,7 @@ def cross_similarity( xsim.eliminate_zeros() if mode == "connectivity": - xsim = xsim.astype(np.bool) + xsim = xsim.astype(bool) elif mode == "affinity": if bandwidth is None: bandwidth = np.nanmedian(xsim.max(axis=1).data) @@ -519,7 +519,7 @@ def recurrence_matrix( rec.eliminate_zeros() if mode == "connectivity": - rec = rec.astype(np.bool) + rec = rec.astype(bool) elif mode == "affinity": if bandwidth is None: bandwidth = np.nanmedian(rec.max(axis=1).data) Index: librosa-0.9.2/tests/test_display.py =================================================================== --- librosa-0.9.2.orig/tests/test_display.py +++ librosa-0.9.2/tests/test_display.py @@ -630,7 +630,7 @@ def test_unknown_axis(S_abs, axis): np.arange(1, 10.0), # strictly positive -np.arange(1, 10.0), # strictly negative np.arange(-3, 4.0), # signed, - np.arange(2, dtype=np.bool), + np.arange(2, dtype=bool), ], ) # binary def test_cmap_robust(data): Index: librosa-0.9.2/tests/test_decompose.py =================================================================== --- librosa-0.9.2.orig/tests/test_decompose.py +++ librosa-0.9.2/tests/test_decompose.py @@ -160,7 +160,7 @@ def test_nn_filter_mean(): X_filtered = librosa.decompose.nn_filter(X) # Normalize the recurrence matrix so dotting computes an average - rec = librosa.util.normalize(rec.astype(np.float), axis=0, norm=1) + rec = librosa.util.normalize(rec.astype(float), axis=0, norm=1) assert np.allclose(X_filtered, X.dot(rec)) @@ -182,7 +182,7 @@ def test_nn_filter_mean_rec(): assert np.allclose(X_filtered[:, i], X[:, i]) # Normalize the recurrence matrix - rec = librosa.util.normalize(rec.astype(np.float), axis=0, norm=1) + rec = librosa.util.normalize(rec.astype(float), axis=0, norm=1) assert np.allclose(X_filtered[:, 3:], (X.dot(rec))[:, 3:]) @@ -197,7 +197,7 @@ def test_nn_filter_mean_rec_sparse(): X_filtered = librosa.decompose.nn_filter(X, rec=rec) # Normalize the recurrence matrix - rec = librosa.util.normalize(rec.toarray().astype(np.float), axis=0, norm=1) + rec = librosa.util.normalize(rec.toarray().astype(float), axis=0, norm=1) assert np.allclose(X_filtered, (X.dot(rec))) Index: librosa-0.9.2/tests/test_dtw.py =================================================================== --- librosa-0.9.2.orig/tests/test_dtw.py +++ librosa-0.9.2/tests/test_dtw.py @@ -272,7 +272,7 @@ def test_dtw_global_inf(): # path-following to (0, 0) # Construct a cost matrix where full alignment is impossible - C = np.zeros((4, 4), dtype=np.float) + C = np.zeros((4, 4), dtype=float) C[-1, -1] = np.inf with pytest.raises(librosa.ParameterError): librosa.sequence.dtw(C=C, subseq=False) @@ -280,7 +280,7 @@ def test_dtw_global_inf(): def test_dtw_subseq_inf(): # Construct a cost matrix where partial alignment is impossible - C = np.zeros((4, 4), dtype=np.float) + C = np.zeros((4, 4), dtype=float) C[-1, :] = np.inf with pytest.raises(librosa.ParameterError): @@ -289,7 +289,7 @@ def test_dtw_subseq_inf(): def test_dtw_subseq_pass(): # Construct a cost matrix where partial alignment is possible - C = np.zeros((4, 4), dtype=np.float) + C = np.zeros((4, 4), dtype=float) C[-1, 2:] = np.inf librosa.sequence.dtw(C=C, subseq=True) Index: librosa-0.9.2/tests/test_effects.py =================================================================== --- librosa-0.9.2.orig/tests/test_effects.py +++ librosa-0.9.2/tests/test_effects.py @@ -122,8 +122,8 @@ def test_pitch_shift_multi(y_multi): def test_remix_mono(align_zeros): # without zc alignment - y = np.asarray([1, 1, -1, -1, 2, 2, -1, -1, 1, 1], dtype=np.float) - y_t = np.asarray([-1, -1, -1, -1, 1, 1, 1, 1, 2, 2], dtype=np.float) + y = np.asarray([1, 1, -1, -1, 2, 2, -1, -1, 1, 1], dtype=float) + y_t = np.asarray([-1, -1, -1, -1, 1, 1, 1, 1, 2, 2], dtype=float) intervals = np.asarray([[2, 4], [6, 8], [0, 2], [8, 10], [4, 6]]) y_out = librosa.effects.remix(y, intervals, align_zeros=align_zeros) @@ -134,8 +134,8 @@ def test_remix_mono(align_zeros): def test_remix_stereo(align_zeros): # without zc alignment - y = np.asarray([1, 1, -1, -1, 2, 2, -1, -1, 1, 1], dtype=np.float) - y_t = np.asarray([-1, -1, -1, -1, 1, 1, 1, 1, 2, 2], dtype=np.float) + y = np.asarray([1, 1, -1, -1, 2, 2, -1, -1, 1, 1], dtype=float) + y_t = np.asarray([-1, -1, -1, -1, 1, 1, 1, 1, 2, 2], dtype=float) y = np.vstack([y, y]) y_t = np.vstack([y_t, y_t]) Index: librosa-0.9.2/tests/test_multichannel.py =================================================================== --- librosa-0.9.2.orig/tests/test_multichannel.py +++ librosa-0.9.2/tests/test_multichannel.py @@ -49,7 +49,7 @@ def tfr_multi(y_multi): "ndim,axis", [(1, 0), (1, -1), (2, 0), (2, 1), (2, -1), (3, 0), (3, 2), (3, -1), (4, 0), (4, 3), (4, -1)] ) def test_sync_multi(aggregate, ndim, axis): - data = np.ones([6] * ndim, dtype=np.float) + data = np.ones([6] * ndim, dtype=float) # Make some slices that don't fill the entire dimension slices = [slice(1, 3), slice(3, 4)] Index: librosa-0.9.2/tests/test_util.py =================================================================== --- librosa-0.9.2.orig/tests/test_util.py +++ librosa-0.9.2/tests/test_util.py @@ -866,7 +866,7 @@ def test_index_to_slice(idx, idx_min, id "ndim,axis", [(1, 0), (1, -1), (2, 0), (2, 1), (2, -1), (3, 0), (3, 2), (3, -1)] ) def test_sync(aggregate, ndim, axis): - data = np.ones([6] * ndim, dtype=np.float) + data = np.ones([6] * ndim, dtype=float) # Make some slices that don't fill the entire dimension slices = [slice(1, 3), slice(3, 4)] Index: librosa-0.9.2/librosa/core/constantq.py =================================================================== --- librosa-0.9.2.orig/librosa/core/constantq.py +++ librosa-0.9.2/librosa/core/constantq.py @@ -287,7 +287,7 @@ def hybrid_cqt( Returns ------- - CQT : np.ndarray [shape=(..., n_bins, t), dtype=np.float] + CQT : np.ndarray [shape=(..., n_bins, t), dtype=float] Constant-Q energy for each frequency at each time. See Also @@ -468,7 +468,7 @@ def pseudo_cqt( Returns ------- - CQT : np.ndarray [shape=(..., n_bins, t), dtype=np.float] + CQT : np.ndarray [shape=(..., n_bins, t), dtype=float] Pseudo Constant-Q energy for each frequency at each time. Notes @@ -622,7 +622,7 @@ def icqt( Returns ------- - y : np.ndarray, [shape=(..., n_samples), dtype=np.float] + y : np.ndarray, [shape=(..., n_samples), dtype=float] Audio time-series reconstructed from the CQT representation. See Also @@ -889,7 +889,7 @@ def vqt( Returns ------- - VQT : np.ndarray [shape=(..., n_bins, t), dtype=np.complex] + VQT : np.ndarray [shape=(..., n_bins, t), dtype=complex] Variable-Q value each frequency at each time. See Also Index: librosa-0.9.2/librosa/core/spectrum.py =================================================================== --- librosa-0.9.2.orig/librosa/core/spectrum.py +++ librosa-0.9.2/librosa/core/spectrum.py @@ -2544,7 +2544,7 @@ def _spectrogram( Returns ------- - S_out : np.ndarray [dtype=np.float] + S_out : np.ndarray [dtype=float] - If ``S`` is provided as input, then ``S_out == S`` - Else, ``S_out = |stft(y, ...)|**power`` n_fft : int > 0 Index: librosa-0.9.2/librosa/util/utils.py =================================================================== --- librosa-0.9.2.orig/librosa/util/utils.py +++ librosa-0.9.2/librosa/util/utils.py @@ -2181,7 +2181,7 @@ def dtype_c2r(d, *, default=np.float32): mapping = { np.dtype(np.complex64): np.float32, np.dtype(np.complex128): np.float64, - np.dtype(complex): np.dtype(np.float).type, + np.dtype(complex): np.dtype(float).type, } # If we're given a real type already, return it Index: librosa-0.9.2/tests/test_convert.py =================================================================== --- librosa-0.9.2.orig/tests/test_convert.py +++ librosa-0.9.2/tests/test_convert.py @@ -531,7 +531,7 @@ def test_blocks_to_frames(blocks, block_ assert np.allclose(frames, block_length * np.asanyarray(blocks)) # Check dtype - assert np.issubdtype(frames.dtype, np.int) + assert np.issubdtype(frames.dtype, int) @pytest.mark.parametrize("blocks", [0, 1, [10, 20]]) @@ -548,7 +548,7 @@ def test_blocks_to_samples(blocks, block assert np.allclose(samples, np.asanyarray(blocks) * hop_length * block_length) # Check dtype - assert np.issubdtype(samples.dtype, np.int) + assert np.issubdtype(samples.dtype, int) @pytest.mark.parametrize("blocks", [0, 1, [10, 20]]) @@ -568,7 +568,7 @@ def test_blocks_to_time(blocks, block_le ) # Check dtype - assert np.issubdtype(times.dtype, np.float) + assert np.issubdtype(times.dtype, float) @pytest.mark.parametrize("abbr", [False, True]) Index: librosa-0.9.2/tests/test_failures.py =================================================================== --- librosa-0.9.2.orig/tests/test_failures.py +++ librosa-0.9.2/tests/test_failures.py @@ -24,7 +24,7 @@ def test_mono_valid_stereo(): @pytest.mark.xfail(raises=librosa.ParameterError) def test_valid_audio_int(): - y = np.zeros(10, dtype=np.int) + y = np.zeros(10, dtype=int) librosa.util.valid_audio(y) Index: librosa-0.9.2/tests/test_core.py =================================================================== --- librosa-0.9.2.orig/tests/test_core.py +++ librosa-0.9.2/tests/test_core.py @@ -1347,7 +1347,7 @@ def test_amplitude_to_db_complex(): x = np.abs(np.random.randn(1000)) + NOISE_FLOOR with warnings.catch_warnings(record=True) as out: - db1 = librosa.amplitude_to_db(x.astype(np.complex), top_db=None) + db1 = librosa.amplitude_to_db(x.astype(complex), top_db=None) assert len(out) > 0 assert "complex" in str(out[0].message).lower() @@ -1883,7 +1883,7 @@ def test_pcen_drc(S_pcen, bias, power): def test_pcen_complex(): - S = np.ones((9, 30), dtype=np.complex) + S = np.ones((9, 30), dtype=complex) Pexp = np.ones((9, 30)) with warnings.catch_warnings(record=True) as out: