sciris/sciris-files

Sciris Files and Versioning

Reference for file I/O and version management. See full tutorial: docs/tutorials/tut_files.ipynb.

If you need more detail, use your MCP tools (Context7 or GitMCP) to look up current Sciris documentation, or consult the other Sciris skills.

Save/Load Any Object

sc.save('my-sim.obj', sim)           # Gzipped pickle (works with any Python object)
sim = sc.load('my-sim.obj')          # Load — methods and class still work
sc.zsave('fast.obj', sim)            # Zstandard compression (slightly faster)
# sc.load() auto-detects compression format

JSON

sc.savejson('data.json', obj)        # Saves JSONifiable parts of any object
data = sc.loadjson('data.json')      # Returns dict (not original object)
data = sc.readjson(json_string)      # Parse JSON from string

YAML (JSON superset with comments)

data = sc.readyaml(yaml_string)      # Parse YAML (supports comments)
sc.saveyaml('config.yaml', data)

File Utilities

sc.getfilelist('*.ipynb')            # List files matching pattern
sc.thispath()                        # Path of current file (use instead of pathlib.Path)
sc.makefilepath('data/out.csv', makedirs=True)  # Ensure path exists
sc.rmpath('file_or_folder')         # Remove file or folder (auto-detects)

Versioning and Reproducibility

md = sc.metadata()                   # Collect all environment metadata
sc.compareversions(np, '>1.0')       # Version comparison (returns True/False)
sc.require('numpy>1.20')            # Warn/raise if requirement not met

Metadata-enhanced figures

sc.savefig('fig.png', comments='My analysis')  # Saves with metadata (use instead of plt.savefig)
md = sc.loadmetadata('fig.png')                # Retrieve metadata later

Metadata-enhanced archives

sc.savearchive('sim.zip', sim, files='script.py', comments='Full run')
sim = sc.loadarchive('sim.zip')      # Restore object
md = sc.loadmetadata('sim.zip')      # Get metadata separately