Basic Usage Examples

This guide provides practical examples of using the AutoClean pipeline for common EEG processing tasks.

Simple Resting State Processing

Here's a basic example of processing resting-state EEG data:

1. Create a configuration file config.yml:

task: resting_eyesopen
eeg_system: biosemi64
input_file: /path/to/your/data.bdf
line_freq: 60.0
resample_freq: 250
ref_type: average

1. Set up environment variables:

export AUTOCLEAN_DIR=/path/to/output
export AUTOCLEAN_CONFIG=config.yml

1. Run the pipeline:

from autoclean_preprocessing import autoclean_pipeline_v2 as autoclean

# Process resting state data
autoclean.process_resting_eyesopen({
    'task': 'resting_eyesopen',
    'eeg_system': 'biosemi64',
    'input_file': '/path/to/your/data.bdf'
})

Processing Multiple Files

Here's how to process multiple EEG files:

import glob
from pathlib import Path
from autoclean_preprocessing import autoclean_pipeline_v2 as autoclean

# Get all BDF files in a directory
files = glob.glob('/path/to/data/*.bdf')

# Process each file
for file in files:
    config = {
        'task': 'resting_eyesopen',
        'eeg_system': 'biosemi64',
        'input_file': str(Path(file).absolute())
    }
    autoclean.process_resting_eyesopen(config)

Custom Preprocessing Settings

Example with custom preprocessing settings:

config = {
    'task': 'resting_eyesopen',
    'eeg_system': 'biosemi64',
    'input_file': '/path/to/data.bdf',

    # Custom preprocessing settings
    'resample_freq': 500,  # Higher sampling rate
    'ref_type': 'average',
    'crop_start': 10,  # Start at 10 seconds
    'crop_end': 310,   # End at 310 seconds

    # Custom artifact rejection settings
    'bad_channel_criteria': {
        'deviation_threshold': 4.0,
        'correlation_threshold': 0.6,
        'line_noise_threshold': 5.0
    },

    # ICA settings
    'ica_settings': {
        'n_components': 30,
        'random_state': 42,
        'method': 'fastica'
    }
}

autoclean.process_resting_eyesopen(config)

Accessing Processing Results

Example of accessing and analyzing processing results:

from autoclean_preprocessing import autoclean_pipeline_v2 as autoclean

# Get the run record
run_id = "your_run_id"
run_record = autoclean.get_run_record(run_id)

# Access processing metadata
metadata = run_record['metadata']

# Print processing statistics
print(f"Bad channels detected: {metadata['bad_channels']}")
print(f"ICA components removed: {metadata['ica_components_removed']}")
print(f"Processing duration: {metadata['processing_duration']} seconds")

Error Handling

Example with proper error handling:

from autoclean_preprocessing import autoclean_pipeline_v2 as autoclean

try:
    config = {
        'task': 'resting_eyesopen',
        'eeg_system': 'biosemi64',
        'input_file': '/path/to/data.bdf'
    }
    autoclean.process_resting_eyesopen(config)
except FileNotFoundError:
    print("Input file not found. Please check the file path.")
except ValueError as e:
    print(f"Configuration error: {str(e)}")
except Exception as e:
    print(f"An unexpected error occurred: {str(e)}")

Customizing Output Reports

Example of generating custom reports:

from autoclean_preprocessing import autoclean_pipeline_v2 as autoclean

# Process the data
config = {
    'task': 'resting_eyesopen',
    'eeg_system': 'biosemi64',
    'input_file': '/path/to/data.bdf',

    # Custom report settings
    'report_settings': {
        'generate_pdf': True,
        'include_psd': True,
        'include_ica': True,
        'include_topomaps': True
    }
}

autoclean.process_resting_eyesopen(config)