evlib.codec.fileformat.hdf5¶

hdf5 formats for various existing datasets.

Functions

`hdf5append`(data, new_arr)	For utility: append an array to hdf5 data.
`load_event_timestamp_hdf5`(path, key_pairs[, ...])	For utility: load only timestamps from the hdf5 data.
`load_hdf5`(path, key_dtype_pairs)	Load all the contents of .hdf5 file at once.
`open_hdf5`(path)	Open .hdf5 file, not to load them at once.

evlib.codec.fileformat.hdf5.load_hdf5(path, key_dtype_pairs)¶

Load all the contents of .hdf5 file at once.

Parameters:

path (str) – Path to the .hdf5 file.
key_dtype_pairs (list of tuple) – A triplet, or list of triplets of (key of the return dictionary, key of the hdf5 file data, data type for numpy).For example, [(“ts”, “raw_events/ts”, np.int32), (“x”, “raw_events/x”, np.int16), …]

Returns:

{key of the return dictionary: np.ndarray}

Return type:

dict

evlib.codec.fileformat.hdf5.open_hdf5(path)¶

Open .hdf5 file, not to load them at once.

evlib.codec.fileformat.hdf5.load_event_timestamp_hdf5(path, key_pairs, dtype=<class 'numpy.int32'>)¶

For utility: load only timestamps from the hdf5 data.

Parameters:

path (str) – Path to the .hdf5 file.
key_pairs (Tuple[str, str]) – The tuple of (key of the return dictionary, key of the hdf5 file data)
dtype (type)

Returns:

dict

Return type:

Dict[str, ndarray]

evlib.codec.fileformat.hdf5.hdf5append(data, new_arr)¶

For utility: append an array to hdf5 data.

Parameters:

Return type:

None