| Name | sygnet JSON |
| Version |
0.0.13
JSON |
| download |
| home_page | None |
| Summary | Synthetic data using Generative Adversarial Networks |
| upload_time | 2023-10-04 16:54:59 |
| maintainer | Artem Nesterov, Maksim Zubok |
| docs_url | None |
| author | None |
| requires_python | >=3.9 |
| license | None |
| keywords |
synthetic data
neural networks
gan
torch
|
| VCS |
 |
| bugtrack_url |
|
| requirements |
No requirements were recorded.
|
| Travis-CI |
No Travis.
|
| coveralls test coverage |
No coveralls.
|
# SyGNet<img src="sygnet.png" alt="SyGNet Mascot" align="right" width="20%" /></a>
## **Sy**nthetic data using **G**enerative Adversarial **Net**works
*Principal Investigator: Dr Thomas Robinson (thomas.robinson@durham.ac.uk)*
*Research team: Artem Nesterov, Maksim Zubok*
**sygnet** ("sig·net") is a Python package for generating
synthetic data within
social science contexts. The **sygnet** algorithm uses cutting-edge advances in deep learning methods to learn the underlying relationships between variables in a dataset. Users can then generate brand-new, synthetic observations that mimic the real data.
### Installation
To install via pip, you can run the following command at the command line:
`pip install sygnet`
**sygnet** requires:
numpy>=1.21
torch>=1.10.0
scikit-learn>=1.0
pandas>=1.4
datetime
tqdm
### Example implementation
You can find a demonstration of **sygnet** under [examples/basic_example](examples/basic_example.ipynb).
### Current version: 0.0.13 (alpha release)
**Alpha release**: You should expect both functionality and pipelines to change (rapidly and without warning). Comments and bug reports are very welcome!
Replacing linear with sigmoid activation functions to facilitate better training, given automatic scaling of data to 0-1 space.
### Previous releases
**0.0.12**
Minor documentation updates including README.
**0.0.11**
* Bug fix in sampling method
**0.0.10**
* Minor patch to allow for conda-forge release
**0.0.9**
* Rewrite of main interface and underlying functions
* Bulding models now structured in terms of hidden "blocks"
* Added self-attention mechanism
**0.0.8**
* Update `tune()` to provide no k-fold cross validation as default
* Update numpy dependency to fix pre-processing bug
**0.0.7**
* Update internal `train_*` functions to return losses and improve logging
* Update `tune()` function
**0.0.6 and 0.0.5**
* Internal changes to improve code efficiency
* Removes `sygnet_` from all submodule names
* Lowers PyTorch requirement to 1.10 for compatability with OpenCE environments
**0.0.4**
* Adds `tune()` function to run hyperparameter tuning
* Adds model saving functionality to `SygnetModel.fit()`
* Fixes various bugs
* Improves documentation
**0.0.3**
* Fixes column ordering issue when using mixed activation layer
* Updates example
**0.0.2**
* Fixes mixed activation bug where final layer wasn't sent to `device`
* Adds `SygnetModel.transform()` alias for `SygnetModel.sample()`
**0.0.1**
Our first release! This version has been lightly tested and the core functionality has been implemented.
Raw data
{
"_id": null,
"home_page": null,
"name": "sygnet",
"maintainer": "Artem Nesterov, Maksim Zubok",
"docs_url": null,
"requires_python": ">=3.9",
"maintainer_email": "Thomas Robinson <t.robinson7@lse.ac.uk>",
"keywords": "synthetic data,neural networks,GAN,torch",
"author": null,
"author_email": "Thomas Robinson <t.robinson7@lse.ac.uk>",
"download_url": "https://files.pythonhosted.org/packages/58/24/3748335d838ba02d1c7ff28b52a715137e8453ffb22550e4e08d16aa91aa/sygnet-0.0.13.tar.gz",
"platform": null,
"description": "# SyGNet<img src=\"sygnet.png\" alt=\"SyGNet Mascot\" align=\"right\" width=\"20%\" /></a>\n\n## **Sy**nthetic data using **G**enerative Adversarial **Net**works\n\n*Principal Investigator: Dr Thomas Robinson (thomas.robinson@durham.ac.uk)*\n\n*Research team: Artem Nesterov, Maksim Zubok*\n\n\n\n**sygnet** (\"sig\u00b7net\") is a Python package for generating \nsynthetic data within \nsocial science contexts. The **sygnet** algorithm uses cutting-edge advances in deep learning methods to learn the underlying relationships between variables in a dataset. Users can then generate brand-new, synthetic observations that mimic the real data.\n\n### Installation\nTo install via pip, you can run the following command at the command line:\n`pip install sygnet`\n\n**sygnet** requires:\n \n numpy>=1.21\n torch>=1.10.0\n scikit-learn>=1.0\n pandas>=1.4\n datetime\n tqdm\n\n### Example implementation\n\nYou can find a demonstration of **sygnet** under [examples/basic_example](examples/basic_example.ipynb).\n\n### Current version: 0.0.13 (alpha release)\n\n**Alpha release**: You should expect both functionality and pipelines to change (rapidly and without warning). Comments and bug reports are very welcome!\n\nReplacing linear with sigmoid activation functions to facilitate better training, given automatic scaling of data to 0-1 space.\n\n### Previous releases\n\n**0.0.12**\n\nMinor documentation updates including README.\n\n**0.0.11**\n\n* Bug fix in sampling method\n\n**0.0.10**\n\n* Minor patch to allow for conda-forge release\n\n**0.0.9**\n\n* Rewrite of main interface and underlying functions\n* Bulding models now structured in terms of hidden \"blocks\"\n* Added self-attention mechanism\n\n**0.0.8**\n\n* Update `tune()` to provide no k-fold cross validation as default\n* Update numpy dependency to fix pre-processing bug\n\n**0.0.7**\n* Update internal `train_*` functions to return losses and improve logging\n* Update `tune()` function\n\n**0.0.6 and 0.0.5**\n* Internal changes to improve code efficiency\n* Removes `sygnet_` from all submodule names\n* Lowers PyTorch requirement to 1.10 for compatability with OpenCE environments\n\n**0.0.4**\n* Adds `tune()` function to run hyperparameter tuning\n* Adds model saving functionality to `SygnetModel.fit()`\n* Fixes various bugs\n* Improves documentation\n\n**0.0.3**\n* Fixes column ordering issue when using mixed activation layer\n* Updates example\n\n**0.0.2**\n* Fixes mixed activation bug where final layer wasn't sent to `device`\n* Adds `SygnetModel.transform()` alias for `SygnetModel.sample()`\n\n**0.0.1**\nOur first release! This version has been lightly tested and the core functionality has been implemented. \n",
"bugtrack_url": null,
"license": null,
"summary": "Synthetic data using Generative Adversarial Networks",
"version": "0.0.13",
"project_urls": {
"Bugs": "https://github.com/tsrobinson/SyGNet/issues",
"Source": "https://github.com/tsrobinson/SyGNet"
},
"split_keywords": [
"synthetic data",
"neural networks",
"gan",
"torch"
],
"urls": [
{
"comment_text": null,
"digests": {
"blake2b_256": "1a12a79ea517e8893e5ecce8e20292c1b7bb9882472e8be3c94351598b26038a",
"md5": "156eea98c9fce1e9b4533b689a44ce94",
"sha256": "baf21420b93c02cdbe5c1530b3f47fafb97e29713e4b09f7a2aac811ae92ea87"
},
"downloads": -1,
"filename": "sygnet-0.0.13-py3-none-any.whl",
"has_sig": false,
"md5_digest": "156eea98c9fce1e9b4533b689a44ce94",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.9",
"size": 32902,
"upload_time": "2023-10-04T16:54:57",
"upload_time_iso_8601": "2023-10-04T16:54:57.699369Z",
"url": "https://files.pythonhosted.org/packages/1a/12/a79ea517e8893e5ecce8e20292c1b7bb9882472e8be3c94351598b26038a/sygnet-0.0.13-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": null,
"digests": {
"blake2b_256": "58243748335d838ba02d1c7ff28b52a715137e8453ffb22550e4e08d16aa91aa",
"md5": "0f71e6f3c8b8314f5e8c451c0e47a885",
"sha256": "4fe32d6d07ab1a1a85b748906164911a1d4ae64ccdbd32c5b786a05fdc71c023"
},
"downloads": -1,
"filename": "sygnet-0.0.13.tar.gz",
"has_sig": false,
"md5_digest": "0f71e6f3c8b8314f5e8c451c0e47a885",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.9",
"size": 5717080,
"upload_time": "2023-10-04T16:54:59",
"upload_time_iso_8601": "2023-10-04T16:54:59.991403Z",
"url": "https://files.pythonhosted.org/packages/58/24/3748335d838ba02d1c7ff28b52a715137e8453ffb22550e4e08d16aa91aa/sygnet-0.0.13.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2023-10-04 16:54:59",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "tsrobinson",
"github_project": "SyGNet",
"travis_ci": false,
"coveralls": false,
"github_actions": true,
"lcname": "sygnet"
}
