Interactive Explorers
scitex-io includes interactive explorers for hierarchical data formats. These let you browse complex files without loading everything into memory.
H5Explorer
Browse HDF5 files interactively:
from scitex_io import H5Explorer
explorer = H5Explorer("experiment.hdf5")
# List all keys at root level
explorer.keys()
# Navigate into groups
explorer["recordings"].keys()
# Load specific datasets
data = explorer["recordings/channel_01"][:]
# Check if a key exists
from scitex_io import has_h5_key
if has_h5_key("experiment.hdf5", "recordings/channel_01"):
data = explorer["recordings/channel_01"][:]
Note
Requires h5py: pip install h5py
ZarrExplorer
Browse Zarr stores interactively:
from scitex_io import ZarrExplorer
explorer = ZarrExplorer("experiment.zarr")
# List all arrays/groups
explorer.keys()
# Load specific arrays
data = explorer["timeseries"][:]
Note
Requires zarr: pip install zarr
Migration: HDF5 to Zarr
Convert HDF5 files to Zarr format for better cloud and parallel access:
from scitex_io.utils import migrate_h5_to_zarr, migrate_h5_to_zarr_batch
# Single file
migrate_h5_to_zarr("data.hdf5", "data.zarr")
# Batch conversion
migrate_h5_to_zarr_batch("data_dir/", "zarr_dir/")