| Name | edge-research-pipeline JSON |
| Version |
0.1.4
JSON |
| download |
| home_page | None |
| Summary | Modular pipeline for quantitative signal discovery and validation |
| upload_time | 2025-08-04 09:22:07 |
| maintainer | None |
| docs_url | None |
| author | None |
| requires_python | >=3.8 |
| license | Edge Research Pipeline — Personal Use License (ERPUL)
Copyright (c) Kalle Fischer
Permission is hereby granted to individuals to use, modify, and explore this software **for personal, non-commercial, and academic learning purposes** free of charge, subject to the following conditions:
---
## 1. Personal and Student Use
* You may use this software for personal projects, learning, and experimentation.
* Students may use this software in academic coursework, theses, and research projects without payment.
## 2. Academic Publication Use
* Academic researchers may use this software in published work **free of charge**, provided that the publication explicitly refers to this project (via citation, footnote, or GitHub link).
* If a citation is not included, a license fee **would normally apply**.
* Academic users should contact the author for confirmation or payment details:
**Email:** [contact@khf-research.com](mailto:contact@khf-research.com)
## 3. Commercial and Professional Use
* Use of this software in a commercial setting (including internal research at a for-profit company, paid consulting, or use in a production environment) requires a paid commercial license.
* You are required to contact the author to disclose commercial use and request payment instructions.
**Email:** [contact@khf-research.com](mailto:contact@khf-research.com)
## 4. Redistribution
* Redistribution of this code or derivatives, whether modified or unmodified, is **not allowed** without written permission.
## 5. Disclaimer
This software is provided "as is", without warranty of any kind, express or implied. Use it at your own risk.
---
## Contact
To notify of commercial use or request academic waiver confirmation, please contact:
**Kalle Fischer**
**Email:** [contact@khf-research.com](mailto:contact@khf-research.com)
**Project URL:** [https://github.com/KHFischer/edge-research-pipeline](https://github.com/KHFischer/edge-research-pipeline)
You may also submit licensing questions or intent to use commercially through the project issue tracker.
---
This license may be updated or replaced in future versions of the project. Any changes will be clearly documented and versioned.
|
| keywords |
rule mining
pattern discovery
interpretable machine learning
feature engineering
tabular data
tabular machine learning
subgroup discovery
signal discovery
data validation
data cleaning pipeline
rule-based modeling
backtesting
bootstrap resampling
walk-forward analysis
synthetic data generation
quantitative research
financial machine learning
quantitative finance
python package
|
| VCS |
 |
| bugtrack_url |
|
| requirements |
badgers
google_auth
imodels
joblib
mlxtend
numpy
orange3
pandas
params
pysubgroup
PyYAML
scikit_eLCS
scikit-learn
scipy
sdv
statsmodels
synthcity
tqdm
|
| Travis-CI |
No Travis.
|
| coveralls test coverage |
No coveralls.
|
# 🧠 Edge Research Pipeline
The **Edge Research Pipeline** is a modular, privacy-first research toolkit for **rule mining**, **pattern discovery**, and **interpretable machine learning** on **tabular datasets**. It supports automated **feature engineering**, **target labeling**, **robust validation**, and **signal discovery** workflows across domains including **quantitative finance**, **structured data mining**, and **subgroup analysis**, its techniques are broadly applicable to any domain involving structured data and statistical rule discovery.
---
## 🚀 Key Features
A flexible, modular Python library enabling you to:
* **Clean, normalize, and transform** tabular datasets
* **Engineer features** relevant to finance, statistics, and other structured-data domains
* **Generate and label custom targets** for supervised tasks
* **Discover signals** using rule mining and pattern search methods
* **Perform robust validation tests** (e.g., train/test splits, bootstrap, walk-forward analysis, false discovery rate)
* **Reproduce results** with complete configuration export and local-only processing
* **Efficiently execute parameter grids** via function calls or a CLI
---
## 🔒 Privacy by Design
All computations run **locally**—no data ever leaves your environment. Designed explicitly for regulated industries, confidential research, and reproducible workflows.
---
## 📦 Installation
Install required dependencies using:
```bash
pip install -r ./requirements.txt
```
**Note:** Dependencies were generated via `pipreqs` and may need further validation.
---
## ⚠️ Compatibility Notes & Optional Dependencies
This project includes optional support for advanced mining and synthetic data tools like `orange3` and `synthcity`. These libraries are powerful but have strict, conflicting version requirements that cannot be satisfied simultaneously in a single install.
### 🧨 Known Conflicts
* `orange3` requires `xgboost >=1.7.4, <2.1`
* `synthcity` requires `xgboost >=2.1.0`
* `xgbse` (a dependency of `synthcity`) enforces this version split
* Installing both libraries together will cause `pip install` to fail due to an irreconcilable conflict on `xgboost`
### ✅ Resolution
To avoid these conflicts:
* The core package **does not include** `orange3` or `synthcity` by default
* You can install them separately using **extras**:
```bash
pip install edge-research-pipeline[orange] # for orange3-based rule data generation
pip install edge-research-pipeline[synth] # for synthetic data workflows
```
⚠️ **Note:** Installing both `orange3` and `synthcity` via extras will fail due to incompatible `xgboost` requirements.
If you need both, install the pipeline without either extra:
```bash
pip install edge-research-pipeline
```
Then manually install each library:
```bash
pip install orange3
pip install synthcity
```
This bypasses pip’s dependency resolver and allows both to coexist — but may require you to manage compatibility manually.
---
### ⚠️ Additional Dependency Warnings
Some third-party tools (e.g., `torch`, `scipy`, `pandas`, `databricks`, `ydata-profiling`) may also have mutually incompatible version constraints depending on your environment. We strongly recommend installing this package in a **clean virtual environment** to prevent dependency resolution issues:
```bash
python -m venv erp_env
.\erp_env\Scripts\activate # Windows
# source erp_env/bin/activate # macOS/Linux
pip install edge-research-pipeline
```
---
## 🧩 Quick Start Example
Run a full pipeline example via the command line:
```bash
python edge_research/pipeline/main.py params/grid_params.yaml
```
Or check the ready-to-run examples in the [`examples/`](./examples/) directory.
---
<!--
Keywords:
rule mining, pattern discovery, interpretable machine learning, feature engineering,
subgroup discovery, tabular ML, signal validation, financial machine learning, data cleaning pipeline,
synthcity, orange3, CN2 rule induction, robust backtesting, rule-based modeling, bootstrapping, walk-forward analysis
-->
## 📁 Project Structure
```text
edge-research-pipeline
├── data/ # Sample datasets (sandbox only)
├── docs/ # Documentation per module
├── edge_research/ # Core logic modules
│ ├── logger/
│ ├── pipeline/
│ ├── preprocessing/
│ ├── rules_mining/
│ ├── statistics/
│ ├── utils/
│ └── validation_tests/
├── examples/ # Copy-pasteable usage examples
├── params/ # Configuration files
├── tests/ # Unit tests for major functions
├── LICENSE
├── README.md
└── requirements.txt
```
Detailed explanations for each subfolder are available within their respective READMEs.
---
## ⚙️ Configuration Philosophy
Configuration files are managed via YAML files within `./params/`:
* **`default_params.yaml`**: Base configuration with mandatory default values (do not modify)
* **`custom_params.yaml`**: Override specific parameters from defaults
* **`grid_params.yaml`**: Parameters specifically for orchestrating grid pipeline runs
**Precedence hierarchy:**
* For pipeline runs (`pipeline.py` or CLI):
`grid_params > custom_params > default_params`
* For direct function calls:
`custom_params > default_params`
Parameters can also be directly overridden by passing a Python dictionary at runtime.
---
## 🧪 Testing
Unit tests cover all major logical functions, ensuring correctness and robustness. Tests are written using `pytest`. Short utility functions, simple wrappers, and internal helpers are generally not included.
Run tests via:
```bash
pytest tests/
```
---
## 🤝 Contributing
We welcome contributions! Follow these guidelines:
* Keep your commits focused and atomic
* Always provide clear, descriptive commit messages
* Add or update tests for any new feature or bug fix
* Follow existing code style (e.g., use `black` and `flake8` for Python formatting)
* Document new functionality thoroughly within the relevant `.md` file in `docs/`
* Respect privacy-by-design principles—no logging or external data exposure
Feel free to open issues for discussions or submit pull requests directly.
---
## 📄 License
This project is licensed under the **Edge Research Personal Use License (ERPUL)**.
The Edge Research Pipeline is free for personal and academic use.
**Commercial use requires a license.**
👉 See [PRICING.md](./PRICING.md) for full license tiers and support options.
- ✅ Free for personal, student, and academic use (with citation)
- 💼 Commercial use requires approval (temporarily waived)
- 🔒 No redistribution without permission
See [`LICENSE`](./LICENSE) for full terms.
Raw data
{
"_id": null,
"home_page": null,
"name": "edge-research-pipeline",
"maintainer": null,
"docs_url": null,
"requires_python": ">=3.8",
"maintainer_email": null,
"keywords": "rule mining, pattern discovery, interpretable machine learning, feature engineering, tabular data, tabular machine learning, subgroup discovery, signal discovery, data validation, data cleaning pipeline, rule-based modeling, backtesting, bootstrap resampling, walk-forward analysis, synthetic data generation, quantitative research, financial machine learning, quantitative finance, python package",
"author": null,
"author_email": "Kalle Fischer <contact@khf-research.com>",
"download_url": "https://files.pythonhosted.org/packages/7b/b2/a8d3db48e69f4d04bb88502cdc62de02cb15f80d21ece0da7c0c39ea4147/edge_research_pipeline-0.1.4.tar.gz",
"platform": null,
"description": "# \ud83e\udde0 Edge Research Pipeline\r\n\r\nThe **Edge Research Pipeline** is a modular, privacy-first research toolkit for **rule mining**, **pattern discovery**, and **interpretable machine learning** on **tabular datasets**. It supports automated **feature engineering**, **target labeling**, **robust validation**, and **signal discovery** workflows across domains including **quantitative finance**, **structured data mining**, and **subgroup analysis**, its techniques are broadly applicable to any domain involving structured data and statistical rule discovery.\r\n\r\n\r\n---\r\n\r\n## \ud83d\ude80 Key Features\r\n\r\nA flexible, modular Python library enabling you to:\r\n\r\n* **Clean, normalize, and transform** tabular datasets\r\n* **Engineer features** relevant to finance, statistics, and other structured-data domains\r\n* **Generate and label custom targets** for supervised tasks\r\n* **Discover signals** using rule mining and pattern search methods\r\n* **Perform robust validation tests** (e.g., train/test splits, bootstrap, walk-forward analysis, false discovery rate)\r\n* **Reproduce results** with complete configuration export and local-only processing\r\n* **Efficiently execute parameter grids** via function calls or a CLI\r\n\r\n---\r\n\r\n## \ud83d\udd12 Privacy by Design\r\n\r\nAll computations run **locally**\u2014no data ever leaves your environment. Designed explicitly for regulated industries, confidential research, and reproducible workflows.\r\n\r\n---\r\n\r\n## \ud83d\udce6 Installation\r\n\r\nInstall required dependencies using:\r\n\r\n```bash\r\npip install -r ./requirements.txt\r\n```\r\n\r\n**Note:** Dependencies were generated via `pipreqs` and may need further validation.\r\n\r\n---\r\n\r\n## \u26a0\ufe0f Compatibility Notes & Optional Dependencies\r\n\r\nThis project includes optional support for advanced mining and synthetic data tools like `orange3` and `synthcity`. These libraries are powerful but have strict, conflicting version requirements that cannot be satisfied simultaneously in a single install.\r\n\r\n### \ud83e\udde8 Known Conflicts\r\n\r\n* `orange3` requires `xgboost >=1.7.4, <2.1`\r\n* `synthcity` requires `xgboost >=2.1.0`\r\n* `xgbse` (a dependency of `synthcity`) enforces this version split\r\n* Installing both libraries together will cause `pip install` to fail due to an irreconcilable conflict on `xgboost`\r\n\r\n\r\n### \u2705 Resolution\r\n\r\nTo avoid these conflicts:\r\n\r\n* The core package **does not include** `orange3` or `synthcity` by default\r\n* You can install them separately using **extras**:\r\n\r\n ```bash\r\n pip install edge-research-pipeline[orange] # for orange3-based rule data generation\r\n pip install edge-research-pipeline[synth] # for synthetic data workflows\r\n ```\r\n\r\n\u26a0\ufe0f **Note:** Installing both `orange3` and `synthcity` via extras will fail due to incompatible `xgboost` requirements.\r\nIf you need both, install the pipeline without either extra:\r\n\r\n```bash\r\npip install edge-research-pipeline\r\n```\r\n\r\nThen manually install each library:\r\n\r\n```bash\r\npip install orange3\r\npip install synthcity\r\n```\r\n\r\nThis bypasses pip\u2019s dependency resolver and allows both to coexist \u2014 but may require you to manage compatibility manually.\r\n\r\n---\r\n\r\n\r\n### \u26a0\ufe0f Additional Dependency Warnings\r\n\r\nSome third-party tools (e.g., `torch`, `scipy`, `pandas`, `databricks`, `ydata-profiling`) may also have mutually incompatible version constraints depending on your environment. We strongly recommend installing this package in a **clean virtual environment** to prevent dependency resolution issues:\r\n\r\n```bash\r\npython -m venv erp_env\r\n.\\erp_env\\Scripts\\activate # Windows\r\n# source erp_env/bin/activate # macOS/Linux\r\npip install edge-research-pipeline\r\n```\r\n\r\n---\r\n\r\n## \ud83e\udde9 Quick Start Example\r\n\r\nRun a full pipeline example via the command line:\r\n\r\n```bash\r\npython edge_research/pipeline/main.py params/grid_params.yaml\r\n```\r\n\r\nOr check the ready-to-run examples in the [`examples/`](./examples/) directory.\r\n\r\n---\r\n<!--\r\nKeywords:\r\nrule mining, pattern discovery, interpretable machine learning, feature engineering,\r\nsubgroup discovery, tabular ML, signal validation, financial machine learning, data cleaning pipeline,\r\nsynthcity, orange3, CN2 rule induction, robust backtesting, rule-based modeling, bootstrapping, walk-forward analysis\r\n-->\r\n\r\n## \ud83d\udcc1 Project Structure\r\n\r\n```text\r\nedge-research-pipeline\r\n\u251c\u2500\u2500 data/ # Sample datasets (sandbox only)\r\n\u251c\u2500\u2500 docs/ # Documentation per module\r\n\u251c\u2500\u2500 edge_research/ # Core logic modules\r\n\u2502 \u251c\u2500\u2500 logger/\r\n\u2502 \u251c\u2500\u2500 pipeline/\r\n\u2502 \u251c\u2500\u2500 preprocessing/\r\n\u2502 \u251c\u2500\u2500 rules_mining/\r\n\u2502 \u251c\u2500\u2500 statistics/\r\n\u2502 \u251c\u2500\u2500 utils/\r\n\u2502 \u2514\u2500\u2500 validation_tests/\r\n\u251c\u2500\u2500 examples/ # Copy-pasteable usage examples\r\n\u251c\u2500\u2500 params/ # Configuration files\r\n\u251c\u2500\u2500 tests/ # Unit tests for major functions\r\n\u251c\u2500\u2500 LICENSE\r\n\u251c\u2500\u2500 README.md\r\n\u2514\u2500\u2500 requirements.txt\r\n```\r\n\r\nDetailed explanations for each subfolder are available within their respective READMEs.\r\n\r\n---\r\n\r\n## \u2699\ufe0f Configuration Philosophy\r\n\r\nConfiguration files are managed via YAML files within `./params/`:\r\n\r\n* **`default_params.yaml`**: Base configuration with mandatory default values (do not modify)\r\n* **`custom_params.yaml`**: Override specific parameters from defaults\r\n* **`grid_params.yaml`**: Parameters specifically for orchestrating grid pipeline runs\r\n\r\n**Precedence hierarchy:**\r\n\r\n* For pipeline runs (`pipeline.py` or CLI):\r\n `grid_params > custom_params > default_params`\r\n* For direct function calls:\r\n `custom_params > default_params`\r\n\r\nParameters can also be directly overridden by passing a Python dictionary at runtime.\r\n\r\n---\r\n\r\n## \ud83e\uddea Testing\r\n\r\nUnit tests cover all major logical functions, ensuring correctness and robustness. Tests are written using `pytest`. Short utility functions, simple wrappers, and internal helpers are generally not included.\r\n\r\nRun tests via:\r\n\r\n```bash\r\npytest tests/\r\n```\r\n\r\n---\r\n\r\n## \ud83e\udd1d Contributing\r\n\r\nWe welcome contributions! Follow these guidelines:\r\n\r\n* Keep your commits focused and atomic\r\n* Always provide clear, descriptive commit messages\r\n* Add or update tests for any new feature or bug fix\r\n* Follow existing code style (e.g., use `black` and `flake8` for Python formatting)\r\n* Document new functionality thoroughly within the relevant `.md` file in `docs/`\r\n* Respect privacy-by-design principles\u2014no logging or external data exposure\r\n\r\nFeel free to open issues for discussions or submit pull requests directly.\r\n\r\n---\r\n\r\n## \ud83d\udcc4 License\r\n\r\nThis project is licensed under the **Edge Research Personal Use License (ERPUL)**.\r\nThe Edge Research Pipeline is free for personal and academic use. \r\n**Commercial use requires a license.**\r\n\r\n\ud83d\udc49 See [PRICING.md](./PRICING.md) for full license tiers and support options.\r\n\r\n- \u2705 Free for personal, student, and academic use (with citation)\r\n- \ud83d\udcbc Commercial use requires approval (temporarily waived)\r\n- \ud83d\udd12 No redistribution without permission\r\n\r\nSee [`LICENSE`](./LICENSE) for full terms.\r\n\r\n\r\n\r\n\r\n",
"bugtrack_url": null,
"license": "Edge Research Pipeline \u2014 Personal Use License (ERPUL)\r\n \r\n Copyright (c) Kalle Fischer\r\n \r\n Permission is hereby granted to individuals to use, modify, and explore this software **for personal, non-commercial, and academic learning purposes** free of charge, subject to the following conditions:\r\n \r\n ---\r\n \r\n ## 1. Personal and Student Use\r\n \r\n * You may use this software for personal projects, learning, and experimentation.\r\n * Students may use this software in academic coursework, theses, and research projects without payment.\r\n \r\n ## 2. Academic Publication Use\r\n \r\n * Academic researchers may use this software in published work **free of charge**, provided that the publication explicitly refers to this project (via citation, footnote, or GitHub link).\r\n * If a citation is not included, a license fee **would normally apply**.\r\n * Academic users should contact the author for confirmation or payment details:\r\n **Email:** [contact@khf-research.com](mailto:contact@khf-research.com)\r\n \r\n ## 3. Commercial and Professional Use\r\n \r\n * Use of this software in a commercial setting (including internal research at a for-profit company, paid consulting, or use in a production environment) requires a paid commercial license.\r\n * You are required to contact the author to disclose commercial use and request payment instructions.\r\n **Email:** [contact@khf-research.com](mailto:contact@khf-research.com)\r\n \r\n ## 4. Redistribution\r\n \r\n * Redistribution of this code or derivatives, whether modified or unmodified, is **not allowed** without written permission.\r\n \r\n ## 5. Disclaimer\r\n \r\n This software is provided \"as is\", without warranty of any kind, express or implied. Use it at your own risk.\r\n \r\n ---\r\n \r\n ## Contact\r\n \r\n To notify of commercial use or request academic waiver confirmation, please contact:\r\n \r\n **Kalle Fischer** \r\n **Email:** [contact@khf-research.com](mailto:contact@khf-research.com) \r\n **Project URL:** [https://github.com/KHFischer/edge-research-pipeline](https://github.com/KHFischer/edge-research-pipeline)\r\n \r\n You may also submit licensing questions or intent to use commercially through the project issue tracker.\r\n \r\n ---\r\n \r\n This license may be updated or replaced in future versions of the project. Any changes will be clearly documented and versioned.\r\n ",
"summary": "Modular pipeline for quantitative signal discovery and validation",
"version": "0.1.4",
"project_urls": {
"Documentation": "https://github.com/KHFischer/edge-research-pipeline/tree/main/docs",
"Homepage": "https://github.com/KHFischer/edge-research-pipeline",
"Issues": "https://github.com/KHFischer/edge-research-pipeline/issues",
"Source": "https://github.com/KHFischer/edge-research-pipeline"
},
"split_keywords": [
"rule mining",
" pattern discovery",
" interpretable machine learning",
" feature engineering",
" tabular data",
" tabular machine learning",
" subgroup discovery",
" signal discovery",
" data validation",
" data cleaning pipeline",
" rule-based modeling",
" backtesting",
" bootstrap resampling",
" walk-forward analysis",
" synthetic data generation",
" quantitative research",
" financial machine learning",
" quantitative finance",
" python package"
],
"urls": [
{
"comment_text": null,
"digests": {
"blake2b_256": "ed2f4cd8a3f8142c96b1d194e6293e253f5961c70af94bff9f2f52eefc6fccf9",
"md5": "0dcf51017e544d40cbd3ba54841d7b25",
"sha256": "d0d0064a14ccf999179921b17f2022f645bfef49426ca37c74ba1220557362df"
},
"downloads": -1,
"filename": "edge_research_pipeline-0.1.4-py3-none-any.whl",
"has_sig": false,
"md5_digest": "0dcf51017e544d40cbd3ba54841d7b25",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.8",
"size": 113167,
"upload_time": "2025-08-04T09:22:06",
"upload_time_iso_8601": "2025-08-04T09:22:06.679848Z",
"url": "https://files.pythonhosted.org/packages/ed/2f/4cd8a3f8142c96b1d194e6293e253f5961c70af94bff9f2f52eefc6fccf9/edge_research_pipeline-0.1.4-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": null,
"digests": {
"blake2b_256": "7bb2a8d3db48e69f4d04bb88502cdc62de02cb15f80d21ece0da7c0c39ea4147",
"md5": "5f42013e4759620c5b08fc16d5d870b5",
"sha256": "34fe0156ebcb249355dcff3c60ea728d02c9b7af147dda64a282d9efbefa4520"
},
"downloads": -1,
"filename": "edge_research_pipeline-0.1.4.tar.gz",
"has_sig": false,
"md5_digest": "5f42013e4759620c5b08fc16d5d870b5",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.8",
"size": 154107,
"upload_time": "2025-08-04T09:22:07",
"upload_time_iso_8601": "2025-08-04T09:22:07.875907Z",
"url": "https://files.pythonhosted.org/packages/7b/b2/a8d3db48e69f4d04bb88502cdc62de02cb15f80d21ece0da7c0c39ea4147/edge_research_pipeline-0.1.4.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2025-08-04 09:22:07",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "KHFischer",
"github_project": "edge-research-pipeline",
"travis_ci": false,
"coveralls": false,
"github_actions": false,
"requirements": [
{
"name": "badgers",
"specs": [
[
">=",
"0.0.10<0.1"
]
]
},
{
"name": "google_auth",
"specs": [
[
">=",
"2.40.3<3.0"
]
]
},
{
"name": "imodels",
"specs": [
[
">=",
"2.0.0<3.0"
]
]
},
{
"name": "joblib",
"specs": [
[
">=",
"1.4.2<2.0"
]
]
},
{
"name": "mlxtend",
"specs": [
[
">=",
"0.23.4<0.24"
]
]
},
{
"name": "numpy",
"specs": [
[
">=",
"1.26.0<2.0"
]
]
},
{
"name": "orange3",
"specs": [
[
">=",
"3.39.0<4.0"
]
]
},
{
"name": "pandas",
"specs": [
[
">=",
"2.3.1<2.4"
]
]
},
{
"name": "params",
"specs": [
[
">=",
"0.9.0<1.0"
]
]
},
{
"name": "pysubgroup",
"specs": [
[
">=",
"0.8.0<0.9"
]
]
},
{
"name": "PyYAML",
"specs": [
[
">=",
"6.0.2<7.0"
]
]
},
{
"name": "scikit_eLCS",
"specs": [
[
">=",
"1.2.4<1.3"
]
]
},
{
"name": "scikit-learn",
"specs": [
[
">=",
"1.7.1<2.0"
]
]
},
{
"name": "scipy",
"specs": [
[
">=",
"1.16.1<2.0"
]
]
},
{
"name": "sdv",
"specs": [
[
">=",
"1.24.1<1.25"
]
]
},
{
"name": "statsmodels",
"specs": [
[
">=",
"0.14.4<0.15"
]
]
},
{
"name": "synthcity",
"specs": [
[
">=",
"0.2.12<0.3"
]
]
},
{
"name": "tqdm",
"specs": [
[
">=",
"4.67.1<5.0"
]
]
}
],
"lcname": "edge-research-pipeline"
}
