PGLearn provides utilities to save/load files in the HDF5 and JSON formats.
HDF5
To load HDF5 files, use the load_h5, which is the same as HDF5.h5read.
h = load_h5("test_file.h5", "/") # load all HDF5 fileTo save HDF5 files to disk, use the save_h5 function. All keys of the dictionary (and sub-dictionaries) must be Strings.
save_h5("myfile.h5", d)When exporting data to an HDF5 file, only the following data types are supported:
String- Number types supported by HDF5.jl
Arrays of those
Numerical data in an unsupported type will be converted to Float64, when possible.
JSON
Use the load_json and save_json functions to load/save data to/from JSON files.
using PGLearn
# Load a dictionary from a JSON file
d = load_json("my_json_file.json")
# Save a dictionary to a JSON file
save_json("my_new_jsonfile.json", d)
save_json("my_pretty_jsonfile.json", d, indent=2) # prettier formattingCompressed JSON files (.json.gz and .json.bz2) are supported automatically
# Load a dictionary from a compressed JSON file
d = load_json("my_json_file.json.gz")
d = load_json("my_json_file.json.bz2")
# Save a dictionary to a compressed JSON file
save_json("my_new_jsonfile.json.gz", d)
save_json("my_new_jsonfile.json.bz2", d)