OECT-Infra

CN: README_CN.md

A comprehensive data processing infrastructure for OECT (Organic Electrochemical Transistor) experiments

Overview

OECT-Infra is an end-to-end platform that transforms raw experimental data into high-performance structured formats, providing standardized feature engineering, visualization, and reporting capabilities for OECT research.

Key Features

• 🔄 Data Conversion: Parallel batch conversion from CSV/JSON to standardized HDF5 format
• 📊 Lazy-Loading API: Efficient access to experimental metadata and measurement data with intelligent caching
• 🔧 Feature Engineering:
  • V1: Extract transfer characteristics (gm, Von, |I|, etc.) in columnar HDF5 format
  • V2: Advanced DAG-based extraction with YAML configs, Parquet storage, and HuggingFace-style API
• 📁 Unified Data Catalog: SQLite-based indexing with bidirectional file↔database sync
• 📈 Visualization: High-performance plotting with animation/video export
• 📄 Automated Reporting: Configurable PowerPoint generation for stability analysis
• 📉 Degradation Analysis: 17+ power law models with multi-metric comparison framework

Installation

pip install oect-infra

Requirements

• Python 3.11 or higher
• Core dependencies: h5py, pandas, numpy, matplotlib, pydantic, scipy, scikit-learn, PyYAML

Quick Start

Using the Unified Interface

from infra.catalog import UnifiedExperimentManager

# Initialize manager
manager = UnifiedExperimentManager('catalog_config.yaml')

# Get an experiment
exp = manager.get_experiment(chip_id="#20250804008", device_id="3")

# Access data
transfer_data = exp.get_transfer_data()
features = exp.get_features(['gm_max_forward', 'Von_forward'])

# Visualization
fig = exp.plot_transfer_evolution()

Using the Command-Line Interface

# Initialize catalog system
catalog init --auto-config

# Scan and index HDF5 files
catalog scan --path data/raw --recursive

# Synchronize data
catalog sync --direction both

# Query experiments
catalog query --chip "#20250804008" --output table

# Extract Features V2
catalog v2 extract-batch --feature-config v2_ml_ready --workers 4

Features V2 Extraction

# Single experiment with V2
exp = manager.get_experiment(chip_id="#20250804008", device_id="3")
result_df = exp.extract_features_v2('v2_transfer_basic', output_format='dataframe')

# Batch extraction
experiments = manager.search(chip_id="#20250804008")
result = manager.batch_extract_features_v2(
    experiments=experiments,
    feature_config='v2_ml_ready',
    save_format='parquet',
    n_workers=4
)

Architecture

Layered Design

Core Foundation (L0)

• csv2hdf: Data conversion
• experiment: Data access
• oect_transfer: Transfer characteristics analysis
• features: Feature storage

Business Application (L1)

• features_version: Feature workflows V1
• features_v2: Feature engineering V2 system
• visualization: Plotting tools

Application Integration (L2)

• catalog: Unified management
• stability_report: Report generation

Data Flow Pipeline

CSV/JSON → csv2hdf → Raw HDF5 → experiment (lazy-loading)
         → [V1] oect_transfer & features_version → Feature HDF5
         → [V2] features_v2 (DAG compute graph) → Feature Parquet
         → catalog (indexing + workflow metadata) → visualization/stability_report

Configuration

OECT-Infra uses YAML configuration files. Create a catalog_config.yaml:

roots:
  raw_data: "data/raw"
  features_v1: "data/features"
  features_v2: "data/features_v2"

database:
  path: "catalog.db"

sync:
  conflict_strategy: "keep_newer"

Documentation

• Complete Documentation
• Package documentation included in the installed package
• See infra/ subdirectory for detailed module documentation

Examples

Check out example notebooks in the source repository:

• Example notebooks and scripts included in package
• Comprehensive API documentation in module docstrings

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

This project is licensed under the MIT License - see the LICENSE file for details.

Citation

If you use OECT-Infra in your research, please cite:

@software{oect_infra,
  author = {lidonghao},
  title = {OECT-Infra: Data Processing Infrastructure for OECT Experiments},
  year = {2025},
  url = {https://github.com/Durian-leader/oect-infra-package}
}

Support

For issues and questions:

• GitHub Issues: https://github.com/Durian-leader/oect-infra-package/issues
• Email: lidonghao100@outlook.com

Acknowledgments

This project was developed for OECT (Organic Electrochemical Transistor) research, providing tools for efficient data management, analysis, and visualization in materials science and electrochemistry research.