ionworks-schema 0.15.2

Pydantic schemas for building Ionworks pipeline configurations

Ionworks Schema

This is a read-only mirror. The source of truth is a private repo.

Pydantic schemas for building Ionworks pipeline configurations.

Overview

Ionworks Schema (ionworks_schema) provides the schema for constructing Ionworks pipeline configurations. Use these classes to define pipelines (data fits, calculations, entries, validations) in Python with validation, then export JSON to submit via the Ionworks API. Pipeline concepts, objectives, and workflows are described in the Pipeline documentation and in the Ionworks documentation.

Pipelines are executed by submitting configurations to the Ionworks API. Use the ionworks-api Python client to create and run jobs: pip install ionworks-api.

Installation

pip install ionworks-schema

Quick start

Build a pipeline configuration with schema classes, export to JSON, and submit with the Ionworks API client:

import ionworks_schema as iws
import json

# Define a parameter to fit (name, initial_value, bounds)
parameter = iws.Parameter(
    name="Positive electrode capacity [A.h]",
    initial_value=1.0,
    bounds=(0.5, 2.0),
)

# Objective: MSMR half-cell fit; data can be "db:<measurement_id>" for uploaded data (use objectives submodule)
objective = iws.objectives.MSMRHalfCell(
    data_input="db:your-measurement-id",
    options={"model": {"electrode": "positive"}},
)

data_fit = iws.DataFit(
    objectives={"ocp": objective},
    parameters={"Positive electrode capacity [A.h]": parameter},
)

pipeline = iws.Pipeline(elements={"fit": data_fit})

# Export to JSON for API submission.
# `to_config()` is the only supported serializer — do NOT use Pydantic's
# `model_dump()`, which is not a wire format (it drops the `type` discriminator
# and skips field-name mappings, producing a dict the API may reject).
config = pipeline.to_config()
with open("pipeline_config.json", "w") as f:
    json.dump(config, f, indent=2)

# Submit via ionworks-api (requires credentials and project ID — see ionworks-api README)
# from ionworks import Ionworks
# client = Ionworks()
# job = client.pipeline.create(config)
# client.pipeline.wait_for_completion(job.id, timeout=600)

Schema classes and pipeline elements

Schema classes mirror the pipeline configuration format consumed by the Ionworks pipeline and API. Runtime behavior and options are documented in the Pipeline user guide and in the ionworkspipeline package (e.g. parsers, data_fits, objectives).

A pipeline is a top-level Pipeline with a dictionary of named elements. Each element has an element_type: entry, data_fit, calculation, or validation.

Role	Schema class	Description
Top-level	Pipeline	Pipeline configuration with named elements.
Entry	DirectEntry	Supply fixed parameter values (no fitting or calculation).
Data fit	DataFit, ArrayDataFit	Fit model parameters to data; contain objectives and parameters.
Calculation	ionworks_schema.calculations	Run calculations (e.g. OCP, diffusivity, geometry). See submodule for available classes.
Objectives	MSMRHalfCell, MSMRFullCell, CurrentDriven, CycleAgeing, CalendarAgeing, EIS, Pulse, Resistance, ElectrodeBalancing, OCPHalfCell, and others	Used inside DataFit.objectives to define what to fit. Import from ionworks_schema.objectives (e.g. iws.objectives.MSMRHalfCell).
Parameters	Parameter	name, initial_value, bounds (and optional prior, etc.). Used in DataFit.parameters; dict key is the parameter name.
Priors	Prior	Used in DataFit.priors. Import from ionworks_schema.priors (e.g. iws.priors.Prior).
Library	Material, Library	Built-in material library for initial parameter values.

Material library

Access built-in materials with validated parameter values for use as initial values or entries:

import ionworks_schema as iws

# List available materials
materials = iws.Library.list_materials()

# Get a specific material (e.g. NMC - Verbrugge 2017)
material = iws.Material.from_library("NMC - Verbrugge 2017")
print(material.parameter_values)

Parameter names and interpretation are described in the Pipeline documentation.

Resources

• Pipeline documentation — workflows, objectives, data fits, and pipeline concepts.
• ionworks-api — submit and manage pipelines (pip install ionworks-api).
• Ionworks documentation — product and platform documentation.
• Changelog — release notes for this package. See the full Ionworks changelog for the platform-wide view.

Note

This package provides configuration schemas only. To run pipelines, export JSON with pipeline.to_config() and submit it via the Ionworks API using the ionworks-api client.

to_config() is the only supported serializer. Pydantic's model_dump() is not a wire format: it omits the type discriminator and does not apply the schema's field-name mappings, so it produces a different dict than to_config() — one the API may reject. Always build API payloads with to_config().