Refactor into NeuroAlign package, add pyproject.toml, and update branding

Author: BitForge95

.gitignore

@@ -18,4 +18,5 @@ Thumbs.db
 *.zarr
 data/
 fake_neuropixels.dat
-*.dat
+*.dat
+*.egg-info/

pipeline/__init__.py

@@ -0,0 +1,12 @@
+"""
+High Performance Neuro data pipeline
+"""
+
+from .base_loader import BaseNeuroLoader
+from .ephys_loader import EphysLoader
+from .video_loader import VideoLoader
+from .nifti_loader import BidsNiftiLoader
+from .synchronizer import DataSynchronizer
+from .exporter import Hdf5Exporter
+
+__version__ = "0.1.0"

cli.py pipeline/cli.pycli.py renamed to pipeline/cli.py

@@ -1,17 +1,26 @@
 import argparse
 import os
+import logging
 import numpy as np
-from pipeline.ephys_loader import EphysLoader
-from pipeline.video_loader import VideoLoader
-from pipeline.synchronizer import DataSynchronizer
-from pipeline.nifti_loader import BidsNiftiLoader
+
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s - [%(levelname)s] - %(name)s: %(message)s",
+    datefmt="%Y-%m-%d %H:%M:%S",
+)
+logger = logging.getLogger("ExpAlignCLI")
+
+from .ephys_loader import EphysLoader
+from .video_loader import VideoLoader
+from .synchronizer import DataSynchronizer
+from .nifti_loader import BidsNiftiLoader
+from .exporter import Hdf5Exporter
 
 
 def main():
     parser = argparse.ArgumentParser(
-        description="ExpAlign: High-Performance Neuro-Data Pipeline"
+        description="NeuroAlign: High-Performance Neuro-Data Pipeline"
     )
-
     parser.add_argument(
         "--ephys",
         type=str,
@@ -24,7 +33,6 @@ def main():
     parser.add_argument(
         "--fmri", type=str, required=False, help="Path to the BIDS NIfTI file"
     )
-
     parser.add_argument(
         "--time", type=float, required=True, help="Timestamp in seconds to align"
     )
@@ -37,18 +45,17 @@ def main():
 
     args = parser.parse_args()
 
-    print("\n" + "=" * 40)
-    print(" EXPERANTO MULTIMODAL CLI PIPELINE")
-    print("=" * 40)
+    logger.info("Starting NeuroAlign Multimodal Pipeline")
 
     ephys_loader = None
     video_loader = None
 
     if args.ephys:
         if not os.path.exists(args.ephys):
-            print(f"Generating dummy ephys data at {args.ephys} for CLI test...")
+            logger.warning(
+                f"Generating dummy ephys data at {args.ephys} for CLI test..."
+            )
             np.random.rand(1000).astype(np.float32).tofile(args.ephys)
-
         ephys_loader = EphysLoader(args.ephys, "CLI_Ephys_Dataset", 30000, 384)
         ephys_loader.load_data(filter_rule=args.filter)
 
@@ -57,15 +64,20 @@ def main():
         video_loader.load_data()
 
     if args.fmri:
-        print("\n--- Initializing BIDS fMRI Pipeline ---")
+        logger.info("Initializing BIDS fMRI Pipeline")
         bids_scan = BidsNiftiLoader(args.fmri, "CLI_fMRI_Dataset")
         bids_scan.load_data(filter_rule=args.filter)
 
     if ephys_loader and video_loader:
         sync_engine = DataSynchronizer(ephys_loader, video_loader)
-        sync_engine.get_data_at_time(args.time)
+        ephys_idx, video_frame = sync_engine.get_data_at_time(args.time)
+
+        exporter = Hdf5Exporter()
+        exporter.export_sync_data(
+            time_sec=args.time, ephys_idx=ephys_idx, video_frame=video_frame
+        )
 
-    print("=" * 40 + "\n")
+    logger.info("Pipeline Execution Finished Successfully.")
 
 
 if __name__ == "__main__":

pipeline/ephys_loader.py

@@ -15,7 +15,7 @@ def __init__(self, file_path, dataset_name, sampling_rate, num_channels):
         # COMPOSITION: We plug the FilterEngine into this loader!
         self.filter_engine = StringFilterEngine()
 
-    def load_data(self, filter_rule=None):  # <-- Add a filter_rule argument
+    def load_data(self, filter_rule=None):
         print(f"[{self.dataset_name}] Initializing Zero-Copy Memory Mapping...")
 
         if not os.path.exists(self.file_path):

pipeline/exporter.py

@@ -0,0 +1,34 @@
+import h5py
+import os
+import logging
+
+
+class Hdf5Exporter:
+    """
+    Serializes synchronized multimodal data chunks into HDF5 format
+    for downstream machine learning pipelines (e.g., PyTorch/TensorFlow).
+    """
+
+    def __init__(self, output_filename="neuro_sync_output.h5"):
+        self.output_filename = output_filename
+        self.logger = logging.getLogger("Hdf5Exporter")
+
+    def export_sync_data(self, time_sec, ephys_idx, video_frame, fmri_tr=None):
+        self.logger.info(
+            f"Exporting synchronized slice at T={time_sec}s to {self.output_filename}"
+        )
+
+        # Open an HDF5 file in write ('w') mode
+        with h5py.File(self.output_filename, "w") as hf:
+            # 1. Create a metadata group
+            meta_group = hf.create_group("metadata")
+            meta_group.create_dataset("sync_time_seconds", data=time_sec)
+
+            # 2. Create the data pointers
+            hf.create_dataset("ephys_sync_index", data=ephys_idx)
+            hf.create_dataset("video_sync_frame", data=video_frame)
+
+            if fmri_tr is not None:
+                hf.create_dataset("fmri_sync_volume", data=fmri_tr)
+
+        self.logger.info("Export complete. Data is ready for downstream AI ingestion.")

pyproject.toml

@@ -0,0 +1,24 @@
+[build-system]
+requires = ["setuptools>=61.0"]
+build-backend = "setuptools.build_meta"
+
+[project]
+name = "neuro-align"
+version = "0.1.0"
+authors = [
+  { name="Charan Sai", email="guttulacharansai@gmail.com" },
+]
+description = "A high-performance, memory-mapped multimodal data pipeline for neuroscience."
+readme = "README.md"
+requires-python = ">=3.9"
+dependencies = [
+    "numpy",
+    "h5py",
+    "nibabel"
+]
+
+[project.scripts]
+neuro-align = "pipeline.cli:main"
+
+[tool.setuptools.packages.find]
+include = ["pipeline*"]

requirements.txt

@@ -1,2 +1,4 @@
 numpy
-pytest
+pytest
+h5py
+nibabel