accusleepy 0.5.0__py3-none-any.whl → 0.7.0__py3-none-any.whl

accusleepy/gui/resources.qrc

@@ -13,4 +13,4 @@
         <file alias="zoom_out.png">icons/zoom_out.png</file>
         <file alias="save.png">icons/save.png</file>
     </qresource>
-</RCC>
+</RCC>

accusleepy/gui/text/main_guide.md

@@ -29,7 +29,8 @@ At this point, you can score the recordings manually.
 5. Score all recordings automatically using the classifier
 
 By default, there are three brain state options: REM, wake, and NREM.
-If you want to change this configuration, click the "Settings" tab.
+If you want to change this configuration, click the "Settings" tab and
+choose "Brain states" from the drop-down menu.
 Note that if you change the configuration, you might be unable to load
 existing labels and calibration data, and you may need to train a new
 classification model.
@@ -45,12 +46,14 @@ To select a file in the primary interface, you can either use the
 associated button, or drag/drop the file into the empty box adjacent
 to the button.
 - Recording file: a .parquet or .csv file containing one
-    column of EEG  data and one column of EMG data.
-    The column names must be eeg and emg.
-- Label file: a .csv file with one column titled brain_state
+    column of EEG data and one column of EMG data.
+    The column names must be **eeg** and **emg**.
+- Label file: a .csv file with one column titled **brain_state**
     with entries that are either the undefined brain state (by default, this is -1)
     or one of the digits in your brain state configuration.
     By default, these are 1-3 where REM = 1, wake = 2, NREM = 3.
+    Optionally, there can be a second column named **confidence_score**
+    containing classification confidence scores between 0 and 1.
 - Calibration data file: required for automated scoring. See Section 4
     for details. These have .csv format.
 - Trained classification model: required for automated scoring. See
@@ -126,15 +129,17 @@ To train a new model on your own data:
     type models, this must be an odd number. In general, about 30
     seconds worth of data is enough.
 4. Choose whether the images used to train the model should be
-    deleted once training is complete. (It's generally best to
-    leave this box checked.)
+    deleted once training is complete. It's generally best to
+    leave this box checked. A (temporary) folder for these files
+    will be created in the same location as the trained model.
 5. Choose whether to create a "default" or "real-time"-type model.
     Note that scoring recordings in the primary interface requires
     a default-type model.
-6. Click "Set training image directory" to select the location
-    where the images used to train the model will be saved. A
-    new directory with an automatically generated name will be
-    created inside the directory you choose.
+6. Choose whether to calibrate the model. This process uses part
+    of the training data to make the model's confidence scores
+    more accurately reflect the probability that the output
+    labels are accurate. If using calibration, choose what percent
+    of the training data to set aside for calibration.
 7. Click "Train classification model" and enter a
     filename for the trained model. Training can take some time.
     The console will display progress updates.
@@ -153,10 +158,11 @@ Instructions for automatic scoring using this interface are below.
 4. If you wish to preserve any existing labels in the label file, and
     only overwrite undefined epochs, check the box labeled
     "Only overwrite undefined epochs".
-5. Set the minimum bout length, in seconds. A typical value could be 5.
+5. Choose whether to save confidence scores to the label files.
+6. Set the minimum bout length, in seconds. A typical value could be 5.
     Following automatic labeling, any brain state bout shorter than this
     duration will be reassigned to the surrounding state (if the states
     on either side of the bout are the same).
-6. Click "Score all automatically" to score all recordings in the
+7. Click "Score all automatically" to score all recordings in the
     recording list. To inspect the results, select a recording
     in the list and click "Score manually".

accusleepy/gui/viewer_window.py

@@ -3,7 +3,7 @@
 ################################################################################
 ## Form generated from reading UI file 'viewer_window.ui'
 ##
-## Created by: Qt User Interface Compiler version 6.8.3
+## Created by: Qt User Interface Compiler version 6.7.3
 ##
 ## WARNING! All changes made in this file will be lost when recompiling UI file!
 ################################################################################
@@ -90,7 +90,7 @@ class Ui_ViewerWindow(object):
         self.top_plot_buttons = QVBoxLayout()
         self.top_plot_buttons.setObjectName("top_plot_buttons")
         self.topcontroltopspacer = QSpacerItem(
-            5, 10, QSizePolicy.Policy.Minimum, QSizePolicy.Policy.Fixed
+            5, 10, QSizePolicy.Policy.Minimum, QSizePolicy.Policy.Preferred
         )
 
         self.top_plot_buttons.addItem(self.topcontroltopspacer)
@@ -142,6 +142,12 @@ class Ui_ViewerWindow(object):
 
         self.zoom_and_brightness = QVBoxLayout()
         self.zoom_and_brightness.setObjectName("zoom_and_brightness")
+        self.verticalSpacer_6 = QSpacerItem(
+            5, 5, QSizePolicy.Policy.Minimum, QSizePolicy.Policy.Preferred
+        )
+
+        self.zoom_and_brightness.addItem(self.verticalSpacer_6)
+
         self.brightness_buttons = QHBoxLayout()
         self.brightness_buttons.setObjectName("brightness_buttons")
         self.specbrighter = QPushButton(ViewerWindow)
@@ -227,6 +233,12 @@ class Ui_ViewerWindow(object):
 
         self.zoom_and_brightness.addLayout(self.zoom_buttons)
 
+        self.verticalSpacer_7 = QSpacerItem(
+            5, 5, QSizePolicy.Policy.Minimum, QSizePolicy.Policy.Preferred
+        )
+
+        self.zoom_and_brightness.addItem(self.verticalSpacer_7)
+
         self.top_plot_buttons.addLayout(self.zoom_and_brightness)
 
         self.topcontrolbottomspacer = QSpacerItem(
@@ -235,11 +247,11 @@ class Ui_ViewerWindow(object):
 
         self.top_plot_buttons.addItem(self.topcontrolbottomspacer)
 
-        self.top_plot_buttons.setStretch(0, 1)
-        self.top_plot_buttons.setStretch(1, 10)
-        self.top_plot_buttons.setStretch(2, 3)
-        self.top_plot_buttons.setStretch(3, 10)
-        self.top_plot_buttons.setStretch(4, 10)
+        self.top_plot_buttons.setStretch(0, 2)
+        self.top_plot_buttons.setStretch(1, 5)
+        self.top_plot_buttons.setStretch(2, 5)
+        self.top_plot_buttons.setStretch(3, 12)
+        self.top_plot_buttons.setStretch(4, 12)
 
         self.all_controls.addLayout(self.top_plot_buttons)
 

accusleepy/gui/viewer_window.ui

@@ -75,14 +75,14 @@
       <enum>QLayout::SizeConstraint::SetDefaultConstraint</enum>
      </property>
      <item>
-      <layout class="QVBoxLayout" name="top_plot_buttons" stretch="1,10,3,10,10">
+      <layout class="QVBoxLayout" name="top_plot_buttons" stretch="2,5,5,12,12">
        <item>
         <spacer name="topcontroltopspacer">
          <property name="orientation">
           <enum>Qt::Orientation::Vertical</enum>
          </property>
          <property name="sizeType">
-          <enum>QSizePolicy::Policy::Fixed</enum>
+          <enum>QSizePolicy::Policy::Preferred</enum>
          </property>
          <property name="sizeHint" stdset="0">
           <size>
@@ -184,6 +184,22 @@
        </item>
        <item>
         <layout class="QVBoxLayout" name="zoom_and_brightness">
+         <item>
+          <spacer name="verticalSpacer_6">
+           <property name="orientation">
+            <enum>Qt::Orientation::Vertical</enum>
+           </property>
+           <property name="sizeType">
+            <enum>QSizePolicy::Policy::Preferred</enum>
+           </property>
+           <property name="sizeHint" stdset="0">
+            <size>
+             <width>5</width>
+             <height>5</height>
+            </size>
+           </property>
+          </spacer>
+         </item>
          <item>
           <layout class="QHBoxLayout" name="brightness_buttons">
            <item>
@@ -361,6 +377,22 @@
            </item>
           </layout>
          </item>
+         <item>
+          <spacer name="verticalSpacer_7">
+           <property name="orientation">
+            <enum>Qt::Orientation::Vertical</enum>
+           </property>
+           <property name="sizeType">
+            <enum>QSizePolicy::Policy::Preferred</enum>
+           </property>
+           <property name="sizeHint" stdset="0">
+            <size>
+             <width>5</width>
+             <height>5</height>
+            </size>
+           </property>
+          </spacer>
+         </item>
         </layout>
        </item>
        <item>

accusleepy/models.py

@@ -10,6 +10,7 @@ from accusleepy.constants import (
     MIN_WINDOW_LEN,
     UPPER_FREQ,
 )
+from accusleepy.temperature_scaling import ModelWithTemperature
 
 # height in pixels of each training image
 IMAGE_HEIGHT = (
@@ -56,20 +57,23 @@ def save_model(
     epochs_per_img: int,
     model_type: str,
     brain_state_set: BrainStateSet,
+    is_calibrated: bool,
 ) -> None:
     """Save classification model and its metadata
 
     :param model: classification model
+    :param filename: filename
     :param epoch_length: epoch length used when training the model
     :param epochs_per_img: number of epochs in each model input
     :param model_type: default or real-time
     :param brain_state_set: set of brain state options
-    :param filename: filename
+    :param is_calibrated: whether the model has been calibrated
     """
     state_dict = model.state_dict()
     state_dict.update({"epoch_length": epoch_length})
     state_dict.update({"epochs_per_img": epochs_per_img})
     state_dict.update({"model_type": model_type})
+    state_dict.update({"is_calibrated": is_calibrated})
     state_dict.update(
         {BRAIN_STATES_KEY: brain_state_set.to_output_dict()[BRAIN_STATES_KEY]}
     )
@@ -90,9 +94,15 @@ def load_model(filename: str) -> tuple[SSANN, int | float, int, str, dict]:
     epoch_length = state_dict.pop("epoch_length")
     epochs_per_img = state_dict.pop("epochs_per_img")
     model_type = state_dict.pop("model_type")
+    if "is_calibrated" in state_dict:
+        is_calibrated = state_dict.pop("is_calibrated")
+    else:
+        is_calibrated = False
     brain_states = state_dict.pop(BRAIN_STATES_KEY)
     n_classes = len([b for b in brain_states if b["is_scored"]])
 
     model = SSANN(n_classes=n_classes)
+    if is_calibrated:
+        model = ModelWithTemperature(model)
     model.load_state_dict(state_dict)
     return model, epoch_length, epochs_per_img, model_type, brain_states

accusleepy/signal_processing.py

@@ -9,6 +9,7 @@ from tqdm import trange
 from accusleepy.brain_state_set import BrainStateSet
 from accusleepy.constants import (
     ANNOTATIONS_FILENAME,
+    CALIBRATION_ANNOTATION_FILENAME,
     DEFAULT_MODEL_TYPE,
     DOWNSAMPLING_START_FREQ,
     EMG_COPIES,
@@ -17,7 +18,7 @@ from accusleepy.constants import (
     MIN_WINDOW_LEN,
     UPPER_FREQ,
 )
-from accusleepy.fileio import Recording, load_labels, load_recording
+from accusleepy.fileio import Recording, load_labels, load_recording, EMGFilter
 from accusleepy.multitaper import spectrogram
 
 # note: scipy is lazily imported
@@ -171,7 +172,10 @@ def create_spectrogram(
 
 
 def get_emg_power(
-    emg: np.array, sampling_rate: int | float, epoch_length: int | float
+    emg: np.array,
+    sampling_rate: int | float,
+    epoch_length: int | float,
+    emg_filter: EMGFilter,
 ) -> np.array:
     """Calculate EMG power for each epoch
 
@@ -181,18 +185,14 @@ def get_emg_power(
     :param emg: EMG signal
     :param sampling_rate: sampling rate, in Hz
     :param epoch_length: epoch length, in seconds
+    :param emg_filter: EMG filter parameters
     :return: EMG "power" for each epoch
     """
     from scipy.signal import butter, filtfilt
 
-    # filter parameters
-    order = 8
-    bp_lower = 20
-    bp_upper = 50
-
     b, a = butter(
-        N=order,
-        Wn=[bp_lower, bp_upper],
+        N=emg_filter.order,
+        Wn=[emg_filter.bp_lower, emg_filter.bp_upper],
         btype="bandpass",
         output="ba",
         fs=sampling_rate,
@@ -215,6 +215,7 @@ def create_eeg_emg_image(
     emg: np.array,
     sampling_rate: int | float,
     epoch_length: int | float,
+    emg_filter: EMGFilter,
 ) -> np.array:
     """Stack EEG spectrogram and EMG power into an image
 
@@ -226,6 +227,7 @@ def create_eeg_emg_image(
     :param emg: EMG signal
     :param sampling_rate: sampling rate, in Hz
     :param epoch_length: epoch length, in seconds
+    :param emg_filter: EMG filter parameters
     :return: combined EEG + EMG image for a recording
     """
     spec, f = create_spectrogram(eeg, sampling_rate, epoch_length)
@@ -241,7 +243,7 @@ def create_eeg_emg_image(
         ]
     )
 
-    emg_log_rms = get_emg_power(emg, sampling_rate, epoch_length)
+    emg_log_rms = get_emg_power(emg, sampling_rate, epoch_length, emg_filter)
     output = np.concatenate(
         [modified_spectrogram, np.tile(emg_log_rms, (EMG_COPIES, 1))]
     )
@@ -369,6 +371,8 @@ def create_training_images(
     epochs_per_img: int,
     brain_state_set: BrainStateSet,
     model_type: str,
+    calibration_fraction: float,
+    emg_filter: EMGFilter,
 ) -> list[int]:
     """Create training dataset
 
@@ -382,6 +386,8 @@ def create_training_images(
     :param epochs_per_img: # number of epochs shown in each image
     :param brain_state_set: set of brain state options
     :param model_type: default or real-time
+    :param calibration_fraction: fraction of training data to use for calibration
+    :param emg_filter: EMG filter parameters
     :return: list of the names of any recordings that could not
             be used to create training images.
     """
@@ -404,9 +410,11 @@ def create_training_images(
                 epoch_length=epoch_length,
             )
 
-            labels = load_labels(recording.label_file)
+            labels, _ = load_labels(recording.label_file)
             labels = brain_state_set.convert_digit_to_class(labels)
-            img = create_eeg_emg_image(eeg, emg, sampling_rate, epoch_length)
+            img = create_eeg_emg_image(
+                eeg, emg, sampling_rate, epoch_length, emg_filter
+            )
             img = mixture_z_score_img(
                 img=img, brain_state_set=brain_state_set, labels=labels
             )
@@ -442,10 +450,25 @@ def create_training_images(
             print(e)
             failed_recordings.append(recording.name)
 
-    # annotation file containing info on all images
-    pd.DataFrame({FILENAME_COL: filenames, LABEL_COL: all_labels}).to_csv(
-        os.path.join(output_path, ANNOTATIONS_FILENAME),
-        index=False,
-    )
+    annotations = pd.DataFrame({FILENAME_COL: filenames, LABEL_COL: all_labels})
+
+    # split into training and calibration sets, if necessary
+    if calibration_fraction > 0:
+        calibration_set = annotations.sample(frac=calibration_fraction)
+        training_set = annotations.drop(calibration_set.index)
+        training_set.to_csv(
+            os.path.join(output_path, ANNOTATIONS_FILENAME),
+            index=False,
+        )
+        calibration_set.to_csv(
+            os.path.join(output_path, CALIBRATION_ANNOTATION_FILENAME),
+            index=False,
+        )
+    else:
+        # annotation file contains info on all training images
+        annotations.to_csv(
+            os.path.join(output_path, ANNOTATIONS_FILENAME),
+            index=False,
+        )
 
     return failed_recordings

accusleepy/temperature_scaling.py

@@ -0,0 +1,157 @@
+import numpy as np
+import torch
+from torch import nn, optim
+from torch.nn import functional as F
+
+
+class ModelWithTemperature(nn.Module):
+    """
+    A thin decorator, which wraps a model with temperature scaling
+    model (nn.Module):
+        A classification neural network
+        NB: Output of the neural network should be the classification logits,
+            NOT the softmax (or log softmax)!
+    """
+
+    def __init__(self, model):
+        super(ModelWithTemperature, self).__init__()
+        self.model = model
+        # https://github.com/gpleiss/temperature_scaling/issues/20
+        # for another approach, see https://github.com/gpleiss/temperature_scaling/issues/36
+        self.model.eval()
+        self.temperature = nn.Parameter(torch.ones(1) * 1.5)
+
+    def forward(self, x):
+        logits = self.model(x)
+        return self.temperature_scale(logits)
+
+    def temperature_scale(self, logits):
+        """
+        Perform temperature scaling on logits
+        """
+        # Expand temperature to match the size of logits
+        temperature = self.temperature.unsqueeze(1).expand(
+            logits.size(0), logits.size(1)
+        )
+        return logits / temperature
+
+    # This function probably should live outside of this class, but whatever
+    def set_temperature(self, valid_loader):
+        """
+        Tune the temperature of the model (using the validation set).
+        We're going to set it to optimize NLL.
+        valid_loader (DataLoader): validation set loader
+        """
+        if torch.accelerator.is_available():
+            device = torch.accelerator.current_accelerator().type
+        else:
+            device = "cpu"
+
+        # self.cuda()
+        self.to(device)
+        nll_criterion = nn.CrossEntropyLoss().to(device)  # .cuda()
+        ece_criterion = _ECELoss().to(device)  # .cuda()
+
+        # First: collect all the logits and labels for the validation set
+        logits_list = []
+        labels_list = []
+        prediction_list = []
+        with torch.no_grad():
+            for x, label in valid_loader:
+                x = x.to(device)  # .cuda()
+                logits = self.model(x)
+                logits_list.append(logits)
+                labels_list.append(label)
+
+                _, pred = torch.max(logits, 1)
+                prediction_list.append(pred)
+            logits = torch.cat(logits_list).to(device)  # .cuda()
+            labels = torch.cat(labels_list).to(device)  # .cuda()
+            predictions = torch.cat(prediction_list).to(device)
+
+        # Calculate NLL and ECE before temperature scaling
+        before_temperature_nll = nll_criterion(logits, labels).item()
+        before_temperature_ece = ece_criterion(logits, labels).item()
+        print(
+            "Before temperature - NLL: %.3f, ECE: %.3f"
+            % (before_temperature_nll, before_temperature_ece)
+        )
+
+        # Next: optimize the temperature w.r.t. NLL
+        # https://github.com/gpleiss/temperature_scaling/issues/34
+        optimizer = optim.LBFGS([self.temperature], lr=0.01, max_iter=100)
+
+        def eval():
+            optimizer.zero_grad()
+            loss = nll_criterion(self.temperature_scale(logits), labels)
+            loss.backward()
+            return loss
+
+        optimizer.step(eval)
+
+        # Calculate NLL and ECE after temperature scaling
+        after_temperature_nll = nll_criterion(
+            self.temperature_scale(logits), labels
+        ).item()
+        after_temperature_ece = ece_criterion(
+            self.temperature_scale(logits), labels
+        ).item()
+        print("Optimal temperature: %.3f" % self.temperature.item())
+        print(
+            "After temperature - NLL: %.3f, ECE: %.3f"
+            % (after_temperature_nll, after_temperature_ece)
+        )
+
+        val_acc = round(
+            100 * np.mean(labels.cpu().numpy() == predictions.cpu().numpy()), 2
+        )
+        print(f"Validation accuracy: {val_acc}%")
+
+        return self
+
+
+class _ECELoss(nn.Module):
+    """
+    Calculates the Expected Calibration Error of a model.
+    (This isn't necessary for temperature scaling, just a cool metric).
+
+    The input to this loss is the logits of a model, NOT the softmax scores.
+
+    This divides the confidence outputs into equally-sized interval bins.
+    In each bin, we compute the confidence gap:
+
+    bin_gap = | avg_confidence_in_bin - accuracy_in_bin |
+
+    We then return a weighted average of the gaps, based on the number
+    of samples in each bin
+
+    See: Naeini, Mahdi Pakdaman, Gregory F. Cooper, and Milos Hauskrecht.
+    "Obtaining Well Calibrated Probabilities Using Bayesian Binning." AAAI.
+    2015.
+    """
+
+    def __init__(self, n_bins=15):
+        """
+        n_bins (int): number of confidence interval bins
+        """
+        super(_ECELoss, self).__init__()
+        bin_boundaries = torch.linspace(0, 1, n_bins + 1)
+        self.bin_lowers = bin_boundaries[:-1]
+        self.bin_uppers = bin_boundaries[1:]
+
+    def forward(self, logits, labels):
+        softmaxes = F.softmax(logits, dim=1)
+        confidences, predictions = torch.max(softmaxes, 1)
+        accuracies = predictions.eq(labels)
+
+        ece = torch.zeros(1, device=logits.device)
+        for bin_lower, bin_upper in zip(self.bin_lowers, self.bin_uppers):
+            # Calculated |confidence - accuracy| in each bin
+            in_bin = confidences.gt(bin_lower.item()) * confidences.le(bin_upper.item())
+            prop_in_bin = in_bin.float().mean()
+            if prop_in_bin.item() > 0:
+                accuracy_in_bin = accuracies[in_bin].float().mean()
+                avg_confidence_in_bin = confidences[in_bin].mean()
+                ece += torch.abs(avg_confidence_in_bin - accuracy_in_bin) * prop_in_bin
+
+        return ece

{accusleepy-0.5.0.dist-info → accusleepy-0.7.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: accusleepy
-Version: 0.5.0
+Version: 0.7.0
 Summary: Python implementation of AccuSleep
 License: GPL-3.0-only
 Author: Zeke Barger
@@ -39,6 +39,7 @@ It offers the following improvements over the MATLAB version (AccuSleep):
     - Model files contain useful metadata (brain state configuration,
       epoch length, number of epochs)
     - Models optimized for real-time scoring can be trained
+- Confidence scores can be saved and visualized
 - Lists of recordings can be imported and exported for repeatable batch processing
 - Undo/redo functionality in the manual scoring interface
 
@@ -75,6 +76,9 @@ to the [config file](accusleepy/config.json).
 
 ## Changelog
 
+- 0.7.0: More settings can be configured in the UI
+- 0.6.0: Confidence scores can now be displayed and saved. Retraining your models is recommended
+    since the new calibration feature will make the confidence scores more accurate.
 - 0.5.0: Performance improvements
 - 0.4.5: Added support for python 3.13, **removed support for python 3.10.**
 - 0.4.4: Performance improvements
@@ -93,7 +97,12 @@ Manual scoring interface
 ## Acknowledgements
 
 We would like to thank [Franz Weber](https://www.med.upenn.edu/weberlab/) for creating an
-early version of the manual labeling interface.
+early version of the manual labeling interface. The code that
+creates spectrograms comes from the
+[Prerau lab](https://github.com/preraulab/multitaper_toolbox/blob/master/python/multitaper_spectrogram_python.py)
+with only minor modifications.
 Jim Bohnslav's [deepethogram](https://github.com/jbohnslav/deepethogram) served as an
 incredibly useful reference when reimplementing this project in python.
+The model calibration code added in version 0.6.0 comes from Geoff Pleiss'
+[temperature scaling repo](https://github.com/gpleiss/temperature_scaling).
 

accusleepy-0.7.0.dist-info/RECORD

@@ -0,0 +1,41 @@
+accusleepy/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+accusleepy/__main__.py,sha256=dKzl2N2Hg9lD264CWYNxThRyDKzWwyMwHRXmJxOmMis,104
+accusleepy/bouts.py,sha256=F_y6DxnpKFfImYb7vCZluZ2eD5I_33gZXmRM8mvebsg,5679
+accusleepy/brain_state_set.py,sha256=fRkrArHLIbEKimub804yt_mUXoyfsjJEfiJnTjeCMkY,3233
+accusleepy/classification.py,sha256=mF35xMrD9QXGldSnl3vkdHbm7CAptPUNjHxUA_agOTA,9778
+accusleepy/config.json,sha256=Ip0qTMAn2LZfof9GVA_azOvpXP0WKnqLCZeSaya1sss,819
+accusleepy/constants.py,sha256=t7x-wzncJ_wVm0Oj6LiUzGukpsTsxfhO0KJiAAsuMN4,2244
+accusleepy/fileio.py,sha256=woIF0zgJt6Lx6T9KBXAQ-AlbQAwOK1_RUmVF710nltI,7383
+accusleepy/gui/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+accusleepy/gui/icons/brightness_down.png,sha256=PLT1fb83RHIhSRuU7MMMx0G7oJAY7o9wUcnqM8veZfM,12432
+accusleepy/gui/icons/brightness_up.png,sha256=64GnUqgPvN5xZ6Um3wOzwqvUmdAWYZT6eFmWpBsHyks,12989
+accusleepy/gui/icons/double_down_arrow.png,sha256=fGiPIP7_RJ3UAonNhORFVX0emXEmtzRlHI3Tfjai064,4964
+accusleepy/gui/icons/double_up_arrow.png,sha256=n7QEo0bZZDve4thwTCKghPKVjTNbQMgyQNsn46iqXbI,5435
+accusleepy/gui/icons/down_arrow.png,sha256=XwS_Gq2j6PoNHRaeaAGoh5kcXJNXWAHWWbJbUsvrRPU,3075
+accusleepy/gui/icons/home.png,sha256=yd3nmHlD9w2a2j3cBd-w_Cuidr-J0apryRoWJoPb66w,5662
+accusleepy/gui/icons/question.png,sha256=IJcIRgQOC9KlzA4vtA5Qu-DQ1-SclhVLeovIsEfl3TU,17105
+accusleepy/gui/icons/save.png,sha256=J3EA8iU1BqLYRSsrq_OdoZlqrv2yfL7oV54DklTy_DI,13555
+accusleepy/gui/icons/up_arrow.png,sha256=V9yF9t1WgjPaUu-mF1YGe_DfaRHg2dUpR_sUVVcvVvY,3329
+accusleepy/gui/icons/zoom_in.png,sha256=MFWnKZp7Rvh4bLPq4Cqo4sB_jQYedUUtT8-ZO8tNYyc,13589
+accusleepy/gui/icons/zoom_out.png,sha256=IB8Jecb3i0U4qjWRR46ridjLpvLCSe7PozBaLqQqYSw,13055
+accusleepy/gui/images/primary_window.png,sha256=ABu49UXTBfI5UwNHrpWPhJoTC7zhyLVvbsbywIuu1nc,602640
+accusleepy/gui/images/viewer_window.png,sha256=b_B7m9WSLMAOzNjctq76SyekO1WfC6qYZVNnYfhjPe8,977197
+accusleepy/gui/images/viewer_window_annotated.png,sha256=uMNUmsZIdzDlQpyoiS3lJGoWlg_T325Oj5hDZhM3Y14,146817
+accusleepy/gui/main.py,sha256=Ywk3XwfwMjAZL76qwDnZRHQ96L7T1DWyMCU8la0N5Qg,62447
+accusleepy/gui/manual_scoring.py,sha256=xk0TERVbH33owj4QYJkMA2LF_qR8jtS5W2enq36_njk,40907
+accusleepy/gui/mplwidget.py,sha256=rJSTtWmLjHn8r3c9Kb23Rc4XzXl3i9B-JrjNjjlNnmQ,13492
+accusleepy/gui/primary_window.py,sha256=sVIsZXnszBsq3ZUC9OIOwB13jPQRcisaxi0Vg-O6A-8,135475
+accusleepy/gui/primary_window.ui,sha256=_SorGY7qJG2B6e45UNt51OheeYLkpzAxxezX1hEfI3Y,201341
+accusleepy/gui/resources.qrc,sha256=wqPendnTLAuKfVI6v2lKHiRqAWM0oaz2ZuF5cucJdS4,803
+accusleepy/gui/resources_rc.py,sha256=Z2e34h30U4snJjnYdZVV9B6yjATKxxfvgTRt5uXtQdo,329727
+accusleepy/gui/text/main_guide.md,sha256=iZDRp5OWyQX9LV7CMeUFIYv2ryKlIcGALRLXjxR8HpI,8288
+accusleepy/gui/text/manual_scoring_guide.md,sha256=ow_RMSjFy05NupEDSCuJtu-V65-BPnIkrZqtssFoZCQ,999
+accusleepy/gui/viewer_window.py,sha256=O4ceqLMYdahxQ9s6DYhloUnNESim-cqIZxFeXEiRjog,24444
+accusleepy/gui/viewer_window.ui,sha256=jsjydsSSyN49AwJw4nVS2mEJ2JBIUTXesAJsij1JNV0,31530
+accusleepy/models.py,sha256=15VjtFoWaYXblyGPbtYgp0yJdyUfGu7t3zCShdtr_7c,3799
+accusleepy/multitaper.py,sha256=D5-iglwkFBRciL5tKSNcunMtcq0rM3zHwRHUVPgem1U,25679
+accusleepy/signal_processing.py,sha256=NOkQuLmUVINd0tFvt48RXRxSl4TDjV42m54XWc_EB9s,16987
+accusleepy/temperature_scaling.py,sha256=glvPcvxHpBdFjwjGfZdNku9L_BozycEmdqZhKKUCCNg,5749
+accusleepy-0.7.0.dist-info/METADATA,sha256=yLkBaWfg1wrN7DevZlLLg881Yq9hhGoTsqHSgCJ0IYk,4536
+accusleepy-0.7.0.dist-info/WHEEL,sha256=b4K_helf-jlQoXBBETfwnf4B04YC67LOev0jo4fX5m8,88
+accusleepy-0.7.0.dist-info/RECORD,,

{accusleepy-0.5.0.dist-info → accusleepy-0.7.0.dist-info}/WHEEL

@@ -1,4 +1,4 @@
 Wheel-Version: 1.0
-Generator: poetry-core 2.1.2
+Generator: poetry-core 2.1.3
 Root-Is-Purelib: true
 Tag: py3-none-any

accusleepy/gui/text/config_guide.txt

@@ -1,29 +0,0 @@
-This is the current brain state configuration.
-If you make changes, click 'Save' to store them.
-
-Each brain state has several attributes:
-
-- Digit: how the brain state is represented in label files,
-    and the key on the keyboard that, during manual scoring,
-    sets an epoch to this brain state.
-
-- Enabled: whether a brain state for this digit exists.
-
-- Name: unique name of the brain state (e.g., REM).
-
-- Scored: whether a classification model should output this
-    brain state. If you have a state that corresponds to
-    missing or corrupted data, for example, you would
-    probably want to uncheck this box.
-
-- Frequency: approximate relative frequency of this brain
-    state. Does not need to be very accurate, but it can
-    influence classification accuracy slightly. The values
-    for all scored brain states must sum to 1.
-
-Important notes:
-- Changing these settings can invalidate existing label files,
-    calibration files, and trained models!
-- Reinstalling AccuSleePy will overwrite this configuration.
-- You can also choose the default epoch length that is shown
-    when the primary interface starts up.