h5_tree()

h5_tree()#

Compact HDF5 Structure Viewer

h5_tree() prints a compact, depth-limited tree of the contents of an open HDF5 file (or group), showing groups and datasets, and (for datasets) the shape and dtype.

It is useful for quickly exploring unfamiliar .hdf5 files without loading large datasets into memory.

Function#

def h5_tree(h5: h5py.File, max_items: int = 500, max_depth: int = 6):
    ...

Parameters#

  • h5 (h5py.File or h5py.Group) An open HDF5 handle (typically returned by h5py.File(…)). You can also pass an HDF5 group (e.g., h5[“/some/group”]) to print a subtree.

  • max_items (int, default: 500) Hard cap on the number of printed entries (groups + datasets). When reached, the traversal stops and prints: … (stopped: max_items reached)

  • max_depth (int, default: 6) Maximum recursion depth to explore.

    • Depth starts at 0 for the root handle you pass in.

    • When depth > max_depth, traversal stops descending further.

Output Format#

Each entry is printed on its own line:

  • Group

    /path/to/group  [Group]

  • Dataset

    /path/to/dataset  [Dataset] shape=(...) dtype=...

At the beginning it prints a header:

Tree (max_items=..., max_depth=...):

Examples#

2) Limit depth (quick high-level overview)#

with h5py.File("myfile.hdf5", "r") as h5:
    h5_tree(h5, max_depth=2)

3) Limit printed items (avoid very large trees)#

with h5py.File("myfile.hdf5", "r") as h5:
    h5_tree(h5, max_items=100)

Notes and Behavior#

  • No data is read from datasets. Only metadata (dataset type, shape, dtype) is accessed.

  • Traversal order follows g.keys() (HDF5 group key order as exposed by h5py).

  • The tree is printed (side effect). The function does not return a value.

Common Use Cases#

  • Discovering dataset paths to use later (e.g., h5[“/path/to/dataset”])

  • Checking where large arrays live and what their shapes/dtypes are

  • Verifying expected file structure in pipelines and batch jobs

Files#

You can find these files in Community Data.

get_now_unix.py
Contents