module adabmDCA.serialization

Portable and atomic serialization helpers for results, logs and archives.

Global Variables

• RESULT_SCHEMA_VERSION

---

function to_jsonable

to_jsonable(value: 'Any') → Any

Recursively convert a value to plain JSON types.

NumPy scalars and arrays, tensors, paths, dataclasses, mappings and sequences are converted; non-finite floats become None.

Args:

• value: Value to convert.

Returns: A structure of dict, list, str, int, float, bool and None only.

---

function result_document

result_document(result_type: 'str', data: 'Mapping[str, Any]') → dict[str, Any]

Wrap result data in a stable, versioned document envelope.

---

function write_json

write_json(
    path: 'str | Path',
    payload: 'Mapping[str, Any]',
    indent: 'int' = 2
) → Path

Write strict UTF-8 JSON atomically.

---

function write_dataframe

write_dataframe(path: 'str | Path', dataframe: 'Any', **kwargs: 'Any') → Path

Write a pandas-compatible dataframe atomically.

---

function write_text

write_text(path: 'str | Path', content: 'str') → Path

Write UTF-8 text atomically.

---

function write_numpy

write_numpy(path: 'str | Path', array: 'ndarray') → Path

Write a NumPy array atomically without altering the requested suffix.

---

function write_numpy_text

write_numpy_text(path: 'str | Path', array: 'ndarray', **kwargs: 'Any') → Path

Write a NumPy array as an atomic text artifact.

---

function resolve_format

resolve_format(
    path: 'str | Path',
    format: 'str | None',
    supported: 'set[str]'
) → str

Resolve an explicit format or infer it from a filename suffix.

---

This file was automatically generated via lazydocs.