eegprep 0.1.0__tar.gz → 0.2.3__tar.gz

{eegprep-0.1.0/src/eegprep.egg-info → eegprep-0.2.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: eegprep
-Version: 0.1.0
+Version: 0.2.3
 Summary: EEG preprocessing pipeline on Python
 Author-email: Arnaud Delorme <adelorme@gmail.com>, Young Truong <dt.young112@gmail.com>
 License: BSD 2-Clause License
@@ -55,10 +55,14 @@ Requires-Dist: python-picard<0.9,>=0.8
 Requires-Dist: scipy>=1.7.0
 Requires-Dist: sympy>=1.14.0
 Requires-Dist: threadpoolctl>=3.6.0
+Requires-Dist: torch
 Provides-Extra: torch
 Requires-Dist: torch>=1.10.0; extra == "torch"
+Provides-Extra: eeglabio
+Requires-Dist: eeglabio>=0.1.2; extra == "eeglabio"
 Provides-Extra: all
 Requires-Dist: torch>=1.10.0; extra == "all"
+Requires-Dist: eeglabio>=0.1.2; extra == "all"
 Dynamic: license-file
 
 # EEGPrep
@@ -96,15 +100,17 @@ The MATLAB and Python implementations were compared using the first two subjects
 
 ## Build Docker
 
-```
-docker run --rm -it -v $(pwd):/usr/src/project dtyoung/eegprep /bin/bash
-docker run -u root --rm -it -v $(pwd):/usr/src/project dtyoung/eegprep /bin/bash
-```
-
-## Remove Docker
+### versioning
+- pyproject version (change version inside file)
+- main version (change version inside file)
+- git tag
+- docker version when building (see below)
 
+### building
 ```
-docker rmi dtyoung/eegprep
+docker build -t eegprep:0.2.3 -f DOCKERFILE .
+docker tag eegprep:0.2.1 arnodelorme/eegprep:0.2.3
+docker push arnodelorme/eegprep:0.2.3  
 ```
 
 Mounted folder in /usr/src/project

{eegprep-0.1.0 → eegprep-0.2.3}/README.md

@@ -33,15 +33,17 @@ The MATLAB and Python implementations were compared using the first two subjects
 
 ## Build Docker
 
-```
-docker run --rm -it -v $(pwd):/usr/src/project dtyoung/eegprep /bin/bash
-docker run -u root --rm -it -v $(pwd):/usr/src/project dtyoung/eegprep /bin/bash
-```
-
-## Remove Docker
+### versioning
+- pyproject version (change version inside file)
+- main version (change version inside file)
+- git tag
+- docker version when building (see below)
 
+### building
 ```
-docker rmi dtyoung/eegprep
+docker build -t eegprep:0.2.3 -f DOCKERFILE .
+docker tag eegprep:0.2.1 arnodelorme/eegprep:0.2.3
+docker push arnodelorme/eegprep:0.2.3  
 ```
 
 Mounted folder in /usr/src/project

{eegprep-0.1.0 → eegprep-0.2.3}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "eegprep"
-version = "0.1.0"
+version = "0.2.3"
 authors = [
   { name="Arnaud Delorme", email="adelorme@gmail.com" },
   { name="Young Truong", email="dt.young112@gmail.com" },
@@ -37,18 +37,23 @@ dependencies = [
   "scipy>=1.7.0",
   "sympy>=1.14.0",
   "threadpoolctl>=3.6.0",
+  "torch"
 ]
 
 [project.optional-dependencies]
 torch = [
   "torch>=1.10.0",
 ]
+eeglabio = [
+  "eeglabio>=0.1.2",
+]
 all = [
   "torch>=1.10.0",
+  "eeglabio>=0.1.2",
 ]
 
 [tool.setuptools]
-include-package-data = false
+include-package-data = true
 
 [tool.setuptools.packages.find]
 where = ["src"]
@@ -58,7 +63,8 @@ exclude = ["eegprep.eeglab*"]
 "eegprep" = [
   "resources/montages/*.locs",
   "resources/montages/README.md",
-  "resources/*.m"
+  "resources/*.m",
+  "netICL.mat"
 ]
 
 [tool.setuptools.exclude-package-data]

{eegprep-0.1.0 → eegprep-0.2.3}/src/eegprep/ICL_feature_extractor.py

@@ -36,7 +36,8 @@ def ICL_feature_extractor(EEG, flag_autocorr=False):
     # Calculate topo
     topo = np.zeros((32, 32, 1, ncomp))
     for it in range(ncomp):
-        _, temp_topo, _, _, _ = topoplot(EEG['icawinv'][:, it], EEG['chanlocs'][EEG['icachansind']], noplot='on')
+        tmp_chanlocs = [EEG['chanlocs'][i] for i in EEG['icachansind']]
+        _, temp_topo, _, _, _ = topoplot(EEG['icawinv'][:, it], tmp_chanlocs, noplot='on')
         temp_topo[np.isnan(temp_topo)] = 0
         topo[:, :, 0, it] = temp_topo / np.max(np.abs(temp_topo))
 

{eegprep-0.1.0 → eegprep-0.2.3}/src/eegprep/clean_artifacts.py

@@ -23,18 +23,18 @@ logger = logging.getLogger(__name__)
 def clean_artifacts(
     EEG: Dict[str, Any],
     # Core parameters
-    ChannelCriterion: Union[float, str] = 0.8,
-    LineNoiseCriterion: Union[float, str] = 4.0,
-    BurstCriterion: Union[float, str] = 5.0,
-    WindowCriterion: Union[float, str] = 0.25,
-    Highpass: Union[str, Tuple[float, float]] = (0.25, 0.75),
+    ChannelCriterion: Union[float, str, None] = 0.8,
+    LineNoiseCriterion: Union[float, str, None] = 4.0,
+    BurstCriterion: Union[float, str, None] = 5.0,
+    WindowCriterion: Union[float, str, None] = 0.25,
+    Highpass: Union[Tuple[float, float], str, None] = (0.25, 0.75),
     # Detail parameters
     ChannelCriterionMaxBadTime: float = 0.5,
-    BurstCriterionRefMaxBadChns: Union[float, str] = 0.075,
-    BurstCriterionRefTolerances: Union[Tuple[float, float], str] = (-np.inf, 5.5),
-    BurstRejection: str = 'off',
-    WindowCriterionTolerances: Union[Tuple[float, float], str] = (-np.inf, 7),
-    FlatlineCriterion: Union[float, str] = 5.0,
+    BurstCriterionRefMaxBadChns: Union[float, str, None] = 0.075,
+    BurstCriterionRefTolerances: Union[Tuple[float, float], str, None] = (-np.inf, 5.5),
+    BurstRejection: bool = False,
+    WindowCriterionTolerances: Union[Tuple[float, float], str, None] = (-np.inf, 7),
+    FlatlineCriterion: Union[float, str, None] = 5.0,
     NumSamples: int = 50,
     SubsetSize: float = 0.25,
     NoLocsChannelCriterion: float = 0.45,
@@ -176,16 +176,16 @@ def clean_artifacts(
     # ------------------------------------------------------------------
     #                     1) Flat‑line channel removal
     # ------------------------------------------------------------------
-    if FlatlineCriterion != 'off':
+    if FlatlineCriterion not in (None, 'off'):
         logger.info('Detecting flat line channels...')
         EEG = clean_flatlines(EEG, max_flatline_duration=float(FlatlineCriterion))
 
     # ------------------------------------------------------------------
     #                        2) High‑pass filtering
     # ------------------------------------------------------------------
-    if Highpass != 'off':
+    if Highpass not in (None, 'off'):
         if not isinstance(Highpass, (tuple, list)) or len(Highpass) != 2:
-            raise ValueError('Highpass must be a (low, high) tuple or "off".')
+            raise ValueError('Highpass must be a (low, high) tuple or None/"off".')
         logger.info('Applying high‑pass filter...')
         EEG = clean_drifts(EEG, tuple(Highpass))
     # Keep a copy after HP for optional return
@@ -195,9 +195,9 @@ def clean_artifacts(
     #            3) Channel cleaning (noisy / disconnected)
     # ------------------------------------------------------------------
     removed_channels = np.zeros(EEG['nbchan'], dtype=bool)
-    if ChannelCriterion != 'off' or LineNoiseCriterion != 'off':
-        chancorr_crit = 0.0 if ChannelCriterion == 'off' else float(ChannelCriterion)
-        line_crit = 100.0 if LineNoiseCriterion == 'off' else float(LineNoiseCriterion)
+    if ChannelCriterion not in (None, 'off') or LineNoiseCriterion not in (None, 'off'):
+        chancorr_crit = 0.0 if ChannelCriterion in (None, 'off') else float(ChannelCriterion)
+        line_crit = 100.0 if LineNoiseCriterion in (None, 'off') else float(LineNoiseCriterion)
         try:
             EEG = clean_channels(
                 EEG,
@@ -226,7 +226,7 @@ def clean_artifacts(
     #                     4) Burst repair via ASR
     # ------------------------------------------------------------------
     BUR = EEG  # default in case ASR is skipped
-    if BurstCriterion != 'off':
+    if BurstCriterion not in (None, 'off'):
         logger.info('Applying ASR burst repair...')
         try:
             BUR = clean_asr(
@@ -250,7 +250,7 @@ def clean_artifacts(
                 maxmem=int(MaxMem),
             )
 
-        if BurstRejection.lower() == 'on':
+        if BurstRejection:
             # Determine unchanged samples after ASR repair
             sample_mask = np.sum(np.abs(EEG['data'] - BUR['data']), axis=0) < 1e-8
             # Convert to intervals (start,end) inclusive
@@ -290,7 +290,7 @@ def clean_artifacts(
     # ------------------------------------------------------------------
     #                     5) Post‑clean windows stage
     # ------------------------------------------------------------------
-    if WindowCriterion != 'off' and WindowCriterionTolerances != 'off':
+    if WindowCriterion not in (None, 'off') and WindowCriterionTolerances not in (None, 'off'):
         logger.info('Final post‑processing – removing irrecoverable windows...')
         EEG, _ = clean_windows(
             EEG,

{eegprep-0.1.0 → eegprep-0.2.3}/src/eegprep/clean_asr.py

@@ -19,9 +19,9 @@ def clean_asr(
     window_len: Optional[float] = None,
     step_size: Optional[int] = None,
     max_dims: float = 0.66,
-    ref_maxbadchannels: Union[float, str, np.ndarray] = 0.075,
-    ref_tolerances: Union[Tuple[float, float], str] = (-3.5, 5.5),
-    ref_wndlen: Union[float, str] = 1.0,
+    ref_maxbadchannels: Union[float, str, None, np.ndarray] = 0.075,
+    ref_tolerances: Union[Tuple[float, float], str, None] = (-3.5, 5.5),
+    ref_wndlen: Union[float, str, None] = 1.0,
     use_gpu: bool = False,
     useriemannian: Optional[str] = None,
     maxmem: Optional[int] = 64
@@ -110,14 +110,14 @@ def clean_asr(
             logger.error(f"An error occurred during clean_windows: {e}")
             logger.warning("Could not automatically identify clean calibration data. Falling back to using the entire data for calibration.")
             ref_section_data = data
-    elif isinstance(ref_maxbadchannels, str) and ref_maxbadchannels.lower() == 'off':
-        logger.info("Using the entire data for calibration ('ref_maxbadchannels' set to 'off').")
+    elif (isinstance(ref_maxbadchannels, str) and ref_maxbadchannels.lower() == 'off') or ref_maxbadchannels is None:
+        logger.info(f"Using the entire data for calibration ('ref_maxbadchannels' set to {ref_maxbadchannels!r}).")
         ref_section_data = data
-    elif isinstance(ref_tolerances, str) and ref_tolerances.lower() == 'off':
-        logger.info("Using the entire data for calibration ('ref_tolerances' set to 'off').")
+    elif (isinstance(ref_tolerances, str) and ref_tolerances.lower() == 'off') or ref_tolerances is None:
+        logger.info(f"Using the entire data for calibration ('ref_tolerances' set to {ref_tolerances!r}).")
         ref_section_data = data
-    elif isinstance(ref_wndlen, str) and ref_wndlen.lower() == 'off':
-        logger.info("Using the entire data for calibration ('ref_wndlen' set to 'off').")
+    elif (isinstance(ref_wndlen, str) and ref_wndlen.lower() == 'off') or ref_wndlen is None:
+        logger.info(f"Using the entire data for calibration ('ref_wndlen' set to {ref_wndlen!r}).")
         ref_section_data = data
     elif isinstance(ref_maxbadchannels, np.ndarray):
         logger.info("Using user-supplied data array for calibration.")
@@ -125,7 +125,7 @@ def clean_asr(
         if ref_section_data.ndim != 2 or ref_section_data.shape[0] != C:
              raise ValueError(f"User-supplied calibration data must be a 2D array with shape ({C}, n_samples).")
     else:
-        raise ValueError(f"Unsupported value or type for 'ref_maxbadchannels': {ref_maxbadchannels}. Must be float, 'off', or numpy array.")
+        raise ValueError(f"Unsupported value or type for 'ref_maxbadchannels': {ref_maxbadchannels}. Must be float, None/'off', or numpy array.")
 
     # --- Calibrate ASR ---
     logger.info('Estimating ASR calibration statistics...')

{eegprep-0.1.0 → eegprep-0.2.3/src/eegprep.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: eegprep
-Version: 0.1.0
+Version: 0.2.3
 Summary: EEG preprocessing pipeline on Python
 Author-email: Arnaud Delorme <adelorme@gmail.com>, Young Truong <dt.young112@gmail.com>
 License: BSD 2-Clause License
@@ -55,10 +55,14 @@ Requires-Dist: python-picard<0.9,>=0.8
 Requires-Dist: scipy>=1.7.0
 Requires-Dist: sympy>=1.14.0
 Requires-Dist: threadpoolctl>=3.6.0
+Requires-Dist: torch
 Provides-Extra: torch
 Requires-Dist: torch>=1.10.0; extra == "torch"
+Provides-Extra: eeglabio
+Requires-Dist: eeglabio>=0.1.2; extra == "eeglabio"
 Provides-Extra: all
 Requires-Dist: torch>=1.10.0; extra == "all"
+Requires-Dist: eeglabio>=0.1.2; extra == "all"
 Dynamic: license-file
 
 # EEGPrep
@@ -96,15 +100,17 @@ The MATLAB and Python implementations were compared using the first two subjects
 
 ## Build Docker
 
-```
-docker run --rm -it -v $(pwd):/usr/src/project dtyoung/eegprep /bin/bash
-docker run -u root --rm -it -v $(pwd):/usr/src/project dtyoung/eegprep /bin/bash
-```
-
-## Remove Docker
+### versioning
+- pyproject version (change version inside file)
+- main version (change version inside file)
+- git tag
+- docker version when building (see below)
 
+### building
 ```
-docker rmi dtyoung/eegprep
+docker build -t eegprep:0.2.3 -f DOCKERFILE .
+docker tag eegprep:0.2.1 arnodelorme/eegprep:0.2.3
+docker push arnodelorme/eegprep:0.2.3  
 ```
 
 Mounted folder in /usr/src/project

{eegprep-0.1.0 → eegprep-0.2.3}/src/eegprep.egg-info/SOURCES.txt

@@ -36,6 +36,7 @@ src/eegprep/epoch.py
 src/eegprep/iclabel.py
 src/eegprep/iclabel_net.py
 src/eegprep/iclabel_net_load_py_measures.py
+src/eegprep/netICL.mat
 src/eegprep/pinv.py
 src/eegprep/pop_epoch.py
 src/eegprep/pop_load_frombids.py

{eegprep-0.1.0 → eegprep-0.2.3}/src/eegprep.egg-info/requires.txt

@@ -12,9 +12,14 @@ python-picard<0.9,>=0.8
 scipy>=1.7.0
 sympy>=1.14.0
 threadpoolctl>=3.6.0
+torch
 
 [all]
 torch>=1.10.0
+eeglabio>=0.1.2
+
+[eeglabio]
+eeglabio>=0.1.2
 
 [torch]
 torch>=1.10.0

{eegprep-0.1.0 → eegprep-0.2.3}/tests/test_clean_artifacts.py

@@ -122,31 +122,31 @@ class TestCleanArtifactsBasic(DebuggableTestCase):
         """Test clean_artifacts with invalid highpass string parameter."""
         with self.assertRaises(ValueError) as cm:
             clean_artifacts(self.test_eeg, Highpass='invalid')
-        self.assertIn('Highpass must be a (low, high) tuple or "off"', str(cm.exception))
+        self.assertIn('Highpass must be a (low, high) tuple or None/"off"', str(cm.exception))
     
     def test_clean_artifacts_invalid_highpass_single_value(self):
         """Test clean_artifacts with single value instead of tuple."""
         with self.assertRaises(ValueError) as cm:
             clean_artifacts(self.test_eeg, Highpass=0.5)
-        self.assertIn('Highpass must be a (low, high) tuple or "off"', str(cm.exception))
+        self.assertIn('Highpass must be a (low, high) tuple or None/"off"', str(cm.exception))
     
     def test_clean_artifacts_invalid_highpass_too_many_values(self):
         """Test clean_artifacts with too many values in highpass tuple."""
         with self.assertRaises(ValueError) as cm:
             clean_artifacts(self.test_eeg, Highpass=(0.1, 0.5, 1.0))
-        self.assertIn('Highpass must be a (low, high) tuple or "off"', str(cm.exception))
+        self.assertIn('Highpass must be a (low, high) tuple or None/"off"', str(cm.exception))
     
     def test_clean_artifacts_invalid_highpass_empty_tuple(self):
         """Test clean_artifacts with empty highpass tuple."""
         with self.assertRaises(ValueError) as cm:
             clean_artifacts(self.test_eeg, Highpass=())
-        self.assertIn('Highpass must be a (low, high) tuple or "off"', str(cm.exception))
+        self.assertIn('Highpass must be a (low, high) tuple or None/"off"', str(cm.exception))
     
     def test_clean_artifacts_invalid_highpass_list_single(self):
         """Test clean_artifacts with single-element list."""
         with self.assertRaises(ValueError) as cm:
             clean_artifacts(self.test_eeg, Highpass=[0.5])
-        self.assertIn('Highpass must be a (low, high) tuple or "off"', str(cm.exception))
+        self.assertIn('Highpass must be a (low, high) tuple or None/"off"', str(cm.exception))
     
     def test_clean_artifacts_valid_highpass_list(self):
         """Test clean_artifacts with valid highpass list (should work like tuple)."""