Fixes for Numpy 2 (#148)

* np.NaN -> np.nan

* remove unneeded 'nyq' argument in firwin

* switch to default_rng

* some more changes

* fix random state copy

* deal with warning

* move to default_rng

* fix imports, no capitalization where not in style

.gitattributes

@@ -1 +1,2 @@
 pyprep/_version.py export-subst
+* text=auto

.github/workflows/python_build.yml

@@ -5,9 +5,6 @@ on:
     branches: [ main ]
   pull_request:
     branches: [ main ]
-  create:
-    branches: [ main ]
-    tags: [ '**' ]
   schedule:
     - cron: "0 4 * * MON"
 
@@ -18,7 +15,7 @@ jobs:
       fail-fast: false
       matrix:
         os: [ubuntu-latest]
-        python-version: ["3.11"]
+        python-version: ["3.12"]
     env:
       TZ: Europe/Berlin
       FORCE_COLOR: true

.github/workflows/python_publish.yml

@@ -20,7 +20,7 @@ jobs:
     - name: Set up Python
       uses: actions/setup-python@v5
       with:
-        python-version: "3.11"
+        python-version: "3.12"
     - name: Install dependencies
       run: |
         python -m pip install --upgrade pip

.github/workflows/python_tests.yml

@@ -5,9 +5,6 @@ on:
     branches: [ main ]
   pull_request:
     branches: [ main ]
-  create:
-    branches: [ main ]
-    tags: [ '**' ]
   schedule:
     - cron: "0 4 * * MON"
 
@@ -17,13 +14,13 @@ jobs:
       fail-fast: false
       matrix:
         platform: [ubuntu-latest, macos-latest, windows-latest]
-        python-version: ["3.11"]
+        python-version: ["3.12"]
         mne-version: [mne-stable]
 
         include:
           # Test mne development version only on ubuntu
           - platform: ubuntu-latest
-            python-version: "3.11"
+            python-version: "3.12"
             mne-version: mne-main
             run-as-extra: true
 

.readthedocs.yml

@@ -8,7 +8,7 @@ version: 2
 build:
   os: ubuntu-22.04
   tools:
-    python: "3.11"
+    python: "3.12"
 
 # Build documentation in the docs/ directory with Sphinx
 sphinx:

README.rst

@@ -16,7 +16,7 @@
 
 
 .. image:: https://readthedocs.org/projects/pyprep/badge/?version=latest
-   :target: http://pyprep.readthedocs.io/en/latest/?badge=latest
+   :target: https://pyprep.readthedocs.io/en/latest/?badge=latest
    :alt: Documentation Status
 
 
@@ -37,8 +37,8 @@ pyprep
 
 For documentation, see the:
 
-- `stable documentation <http://pyprep.readthedocs.io/en/stable/>`_
-- `latest (development) documentation <http://pyprep.readthedocs.io/en/latest/>`_
+- `stable documentation <https://pyprep.readthedocs.io/en/stable/>`_
+- `latest (development) documentation <https://pyprep.readthedocs.io/en/latest/>`_
 
 .. docs_readme_include_label
 

docs/whats_new.rst

@@ -219,7 +219,7 @@ Changelog
 - Initial commit: 2018-04-12
 - Miscellaneous changes
 
-.. _Stefan Appelhoff: http://stefanappelhoff.com/
+.. _Stefan Appelhoff: https://stefanappelhoff.com/
 .. _Aamna Lawrence: https://github.com/AamnaLawrence
 .. _Adam Li: https://github.com/adam2392/
 .. _Christian O'Reilly: https://github.com/christian-oreilly

pyprep/find_noisy_channels.py

@@ -1,6 +1,4 @@
 """finds bad channels."""
-from copy import copy
-
 import mne
 import numpy as np
 from mne.utils import check_random_state, logger
@@ -579,7 +577,14 @@ def find_bad_by_ransac(
         exclude_from_ransac = (
             self.bad_by_correlation + self.bad_by_deviation + self.bad_by_dropout
         )
-        rng = copy(self.random_state) if self.matlab_strict else self.random_state
+
+        if self.matlab_strict:
+            random_state = self.random_state.get_state()
+            rng = np.random.RandomState()
+            rng.set_state(random_state)
+        else:
+            rng = self.random_state
+
         self.bad_by_ransac, ch_correlations_usable = find_bad_by_ransac(
             self.EEGFiltered,
             self.sample_rate,

pyprep/removeTrend.py

@@ -170,5 +170,5 @@ def runline(y, n, dn):
         (np.multiply(np.arange(n + 1, n + npts + 1), a) + b), (npts, 1)
     )
     for i in range(0, len(y_line)):
-        y[i] = y[i] - y_line[i]
+        y[i] = y[i] - y_line[i, 0]
     return y

pyprep/utils.py

@@ -76,7 +76,7 @@ def _mat_quantile(arr, q, axis=None):
     # Sort the array in ascending order along the given axis (any NaNs go to the end)
     # Return NaN if array is empty.
     if len(arr) == 0:
-        return np.NaN
+        return np.nan
     arr_sorted = np.sort(arr, axis=axis)
 
     # Ensure array is either 1D or 2D
@@ -182,7 +182,7 @@ def _eeglab_create_highpass(cutoff, srate):
     N = order + 1
     filt = np.zeros(N)
     filt[N // 2] = 1
-    filt -= firwin(N, transition, window="hamming", nyq=1)
+    filt -= firwin(N, transition, window="hamming")
     return filt
 
 
@@ -215,7 +215,7 @@ def _eeglab_fir_filter(data, filt):
     pad_len = min(group_delay, n_samples)
 
     # Prepare initial state of filter, using padding at start of data
-    start_pad_idx = np.zeros(pad_len, dtype=np.uint8)
+    start_pad_idx = np.zeros(pad_len, dtype=np.uint)
     start_padded = np.concatenate((data[:, start_pad_idx], data[:, :pad_len]), axis=1)
     zi_init = lfilter_zi(filt, 1) * np.take(start_padded, [0], axis=0)
     _, zi = lfilter(filt, 1, start_padded, axis=1, zi=zi_init)
@@ -232,7 +232,7 @@ def _eeglab_fir_filter(data, filt):
         )
 
     # Finish filtering data, using padding at end to calculate final values
-    end_pad_idx = np.zeros(pad_len, dtype=np.uint8) + (n_samples - 1)
+    end_pad_idx = np.zeros(pad_len, dtype=np.uint) + (n_samples - 1)
     end, _ = lfilter(filt, 1, data[:, end_pad_idx], axis=1, zi=zi)
     out[:, (n_samples - pad_len) :] = end[:, (group_delay - pad_len) :]
 

tests/conftest.py

@@ -77,7 +77,7 @@ def make_random_mne_object(
     n_freq_comps=5,
     freq_range=[10, 60],
     scale=1e-6,
-    RNG=np.random.RandomState(1337),
+    rng=np.random.default_rng(1337),
 ):
     """Make a random MNE object to use for testing.
 
@@ -98,8 +98,9 @@ def make_random_mne_object(
     scale : float
         Scaling factor applied to the signal in volts. For example 1e-6 to
         get microvolts.
-    RNG : np.random.RandomState
-        Random state seed.
+    rng : np.random.Generator
+        The random number generator object. Must be created with
+        ``np.random.default_rng``.
 
     Returns
     -------
@@ -120,8 +121,8 @@ def make_random_mne_object(
     high = freq_range[1]
     for chan in range(n_chans):
         # Each channel signal is a sum of random freq sine waves
-        for freq_i in range(n_freq_comps):
-            freq = RNG.randint(low, high, signal_len)
+        for _ in range(n_freq_comps):
+            freq = rng.integers(low, high, signal_len)
             signal[chan, :] += np.sin(2 * np.pi * times * freq)
 
     signal *= scale  # scale

tests/test_find_noisy_channels.py

@@ -1,15 +1,14 @@
 """Test the find_noisy_channels module."""
 import numpy as np
 import pytest
-from numpy.random import RandomState
 
 from pyprep.find_noisy_channels import NoisyChannels
 from pyprep.ransac import find_bad_by_ransac
 from pyprep.removeTrend import removeTrend
 
 # Set a fixed random seed for reproducible test results
 
-RNG = RandomState(30)
+rng = np.random.default_rng(30)
 
 
 # Define some fixtures and utility functions for use across multiple tests
@@ -47,7 +46,7 @@ def raw_tmp(raw_clean_detrend):
 def _generate_signal(fmin, fmax, timepoints, fcount=1):
     """Generate an EEG signal from one or more sine waves in a frequency range."""
     signal = np.zeros_like(timepoints)
-    for freq in RNG.randint(fmin, fmax + 1, fcount):
+    for freq in rng.integers(fmin, fmax + 1, fcount):
         signal += np.sin(2 * np.pi * timepoints * freq)
     return signal * 1e-6
 
@@ -59,7 +58,7 @@ def test_bad_by_nan(raw_tmp):
     """Test the detection of channels containing any NaN values."""
     # Insert a NaN value into a random channel
     n_chans = raw_tmp.get_data().shape[0]
-    nan_idx = int(RNG.randint(0, n_chans, 1))
+    nan_idx = int(rng.integers(0, n_chans, 1)[0])
     raw_tmp._data[nan_idx, 3] = np.nan
 
     # Test automatic detection of NaN channels on NoisyChannels init
@@ -75,7 +74,7 @@ def test_bad_by_flat(raw_tmp):
     """Test the detection of channels with flat or very weak signals."""
     # Make the signal for a random channel extremely weak
     n_chans = raw_tmp.get_data().shape[0]
-    flat_idx = int(RNG.randint(0, n_chans, 1))
+    flat_idx = int(rng.integers(0, n_chans, 1)[0])
     raw_tmp._data[flat_idx, :] = raw_tmp.get_data()[flat_idx, :] * 1e-12
 
     # Test automatic detection of flat channels on NoisyChannels init
@@ -100,7 +99,7 @@ def test_bad_by_deviation(raw_tmp):
 
     # Make the signal for a random channel have a very high amplitude
     n_chans = raw_tmp.get_data().shape[0]
-    high_dev_idx = int(RNG.randint(0, n_chans, 1))
+    high_dev_idx = int(rng.integers(0, n_chans, 1)[0])
     raw_tmp._data[high_dev_idx, :] *= high_dev_factor
 
     # Test detection of abnormally high-amplitude channels
@@ -126,7 +125,7 @@ def test_bad_by_hf_noise(raw_tmp):
     """Test detection of channels with high-frequency noise."""
     # Add some noise between 70 & 80 Hz to the signal of a random channel
     n_chans = raw_tmp.get_data().shape[0]
-    hf_noise_idx = int(RNG.randint(0, n_chans, 1))
+    hf_noise_idx = int(rng.integers(0, n_chans, 1)[0])
     hf_noise = _generate_signal(70, 80, raw_tmp.times, 5) * 10
     raw_tmp._data[hf_noise_idx, :] += hf_noise
 
@@ -148,7 +147,7 @@ def test_bad_by_dropout(raw_tmp):
     """Test detection of channels with excessive portions of flat signal."""
     # Add large dropout portions to the signal of a random channel
     n_chans, n_samples = raw_tmp.get_data().shape
-    dropout_idx = int(RNG.randint(0, n_chans, 1))
+    dropout_idx = int(rng.integers(0, n_chans, 1)[0])
     x1, x2 = (int(n_samples / 10), int(2 * n_samples / 10))
     raw_tmp._data[dropout_idx, x1:x2] = 0  # flatten 10% of signal
 
@@ -162,7 +161,7 @@ def test_bad_by_correlation(raw_tmp):
     """Test detection of channels that correlate poorly with others."""
     # Replace a random channel's signal with uncorrelated values
     n_chans, n_samples = raw_tmp.get_data().shape
-    low_corr_idx = int(RNG.randint(0, n_chans, 1))
+    low_corr_idx = int(rng.integers(0, n_chans, 1)[0])
     raw_tmp._data[low_corr_idx, :] = _generate_signal(10, 30, raw_tmp.times, 5)
 
     # Test detection of channels that correlate poorly with others
@@ -187,7 +186,7 @@ def test_bad_by_SNR(raw_tmp):
     """Test detection of channels that have low signal-to-noise ratios."""
     # Replace a random channel's signal with uncorrelated values
     n_chans = raw_tmp.get_data().shape[0]
-    low_snr_idx = int(RNG.randint(0, n_chans, 1))
+    low_snr_idx = int(rng.integers(0, n_chans, 1)[0])
     raw_tmp._data[low_snr_idx, :] = _generate_signal(10, 30, raw_tmp.times, 5)
 
     # Add some high-frequency noise to the uncorrelated channel
@@ -203,7 +202,7 @@ def test_bad_by_SNR(raw_tmp):
 def test_find_bad_by_ransac(raw_tmp):
     """Test the RANSAC component of NoisyChannels."""
     # Set a consistent random seed for all RANSAC runs
-    RANSAC_RNG = 435656
+    ransac_rng = 435656
 
     # RANSAC identifies channels that go bad together and are highly correlated.
     # Inserting highly correlated signal in channels 0 through 6 at 30 Hz
@@ -222,7 +221,7 @@ def test_find_bad_by_ransac(raw_tmp):
     corr = {}
     for name, args in test_matrix.items():
         nd = NoisyChannels(
-            raw_tmp, do_detrend=False, random_state=RANSAC_RNG, matlab_strict=args[0]
+            raw_tmp, do_detrend=False, random_state=ransac_rng, matlab_strict=args[0]
         )
         nd.find_bad_by_ransac(channel_wise=args[1], max_chunk_size=args[2])
         # Save bad channels and RANSAC correlation matrix for later comparison
@@ -247,7 +246,7 @@ def test_find_bad_by_ransac(raw_tmp):
     assert not np.allclose(corr["by_window"], corr["by_window_strict"])
 
     # Ensure that RANSAC doesn't change random state if in MATLAB-strict mode
-    rng = RandomState(RANSAC_RNG)
+    rng = np.random.RandomState(ransac_rng)
     init_state = rng.get_state()[2]
     nd = NoisyChannels(raw_tmp, do_detrend=False, random_state=rng, matlab_strict=True)
     nd.find_bad_by_ransac()

tests/test_prep_pipeline.py

@@ -223,7 +223,7 @@ def test_prep_pipeline_non_eeg(raw, montage):
         ch_types_non_eeg,
         times,
         sfreq,
-        RNG=np.random.RandomState(1337),
+        rng=np.random.default_rng(1337),
     )
 
     raw_copy.add_channels([raw_non_eeg], force_update_info=True)

tests/test_utils.py

@@ -1,7 +1,6 @@
 """Test various helper functions."""
 import numpy as np
 import pytest
-from numpy.random import RandomState
 
 from pyprep.utils import (
     _correlate_arrays,
@@ -64,7 +63,7 @@ def test_mat_quantile_iqr():
 
     # Add NaNs to test data
     tst_nan = tst.copy()
-    tst_nan[0, :] = np.NaN
+    tst_nan[0, :] = np.nan
 
     # Create arrays containing MATLAB results for NaN test case
     quantile_expected = np.asarray([0.9712, 0.9880, 0.9807])
@@ -91,7 +90,7 @@ def test_mat_quantile_iqr():
 def test_get_random_subset():
     """Test the function for getting random channel subsets."""
     # Generate test data
-    rng = RandomState(435656)
+    rng = np.random.RandomState(435656)
     chans = range(1, 61)
 
     # Compare random subset equivalence with MATLAB