main.py

import numpy as np
import os
from pipeline.ephys_loader import EphysLoader
from pipeline.video_loader import VideoLoader
from pipeline.synchronizer import DataSynchronizer

print("--- GENERATING FAKE NEUROPIXELS DATA ---")
fake_file_path = "fake_neuropixels.dat"
# Generate 1000 random floating-point numbers (simulating brain signals)
fake_data = np.random.rand(1000).astype(np.float32)
fake_data.tofile(fake_file_path)  # Save directly to disk
print(f"Created {fake_file_path} ({os.path.getsize(fake_file_path)} bytes)\n")


# 1. Instantiate loaders (pointing EphysLoader to our new real file)
allen_ephys = EphysLoader(
    file_path=fake_file_path,
    dataset_name="Neuropixels_Probe_A",
    sampling_rate=30000,
    num_channels=384,
)

behavior_cam = VideoLoader(
    file_path="/data/mouse_treadmill.mp4",  # We leave this fake for now
    dataset_name="Treadmill_Camera",
    fps=60,
    resolution="1920x1080",
)

experiment_datasets = [allen_ephys, behavior_cam]
for dataset in experiment_datasets:
    if isinstance(dataset, EphysLoader):
        dataset.load_data(filter_rule="signal > 0.5")
    else:
        dataset.load_data()

# We create the Synchronizer and give it our two loaded objects
sync_engine = DataSynchronizer(ephys_loader=allen_ephys, video_loader=behavior_cam)

# 3. Simulate an event
event_time = 1.5
aligned_ephys_idx, aligned_video_frame = sync_engine.get_data_at_time(event_time)

print("\n--- PIPELINE COMPLETE ---")