# Psi-score
$\psi$-score: Metric of user influence in Online Social Networks
## Requirements
* Python >=3.9,<3.11
## Installation
```bash
$ pip install psi-score
```
## Usage
```python
>>> from psi_score import PsiScore
>>> adjacency = {0: [1, 3], 1: [0, 2], 2: [0, 1, 3], 3: [0]}
>>> lambdas = [0.23, 0.50, 0.86, 0.19]
>>> mus = [0.42, 0.17, 0.10, 0.37]
>>> psiscore = PsiScore()
>>> scores = psiscore.fit_transform(adjacency, lambdas, mus)
>>> scores
array([0.21158803, 0.35253745, 0.28798439, 0.14789014])
>>> np.round(scores, 2)
array([0.21, 0.35, 0.29, 0.15])
```
You can use another algorithm and change some parameters:
```python
>>> psiscore = PsiScore(solver='power_nf', n_iter=500, tol=1e-3)
>>> scores = psiscore.fit_transform(adjacency, lambdas, mus, ps=[1], qs=[0, 3])
```
The ``ps`` and ``qs`` parameters allows to have some chosen ``p_i`` and ``q_i`` vectors (only with the ``push`` and ``power_nf`` methods):
```python
>>> psiscore.P
{1: array([0.5333334 , 0.1681094 , 0.46801851, 0.34442264])}
>>> psiscore.Q
{0: array([0.46164044, 0.0514935 , 0.02798624, 0.30484491]),
3: array([0.13087053, 0.01616898, 0.01850541, 0.42554885])}
```
## License
`psi-score` was created by Nouamane Arhachoui. It is licensed under the terms of the MIT license.
## References
* Giovanidis, A., Baynat, B., Magnien, C., & Vendeville, A. (2021).
Ranking Online Social Users by Their Influence.
IEEE/ACM Transactions on Networking, 29(5), 2198-2214. https://doi.org/10.1109/tnet.2021.3085201
* Arhachoui, N., Bautista, E., Danisch, M., & Giovanidis, A. (2022).
A Fast Algorithm for Ranking Users by their Influence in Online Social Platforms.
2022 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM), 526-533.
https://doi.org/10.1109/ASONAM55673.2022.10068673
Raw data
{
"_id": null,
"home_page": "https://github.com/NouamaneA/psi-score",
"name": "psi-score",
"maintainer": "",
"docs_url": null,
"requires_python": ">=3.9,<3.11",
"maintainer_email": "",
"keywords": "graph algorithms,network science,markov chain,social networks,pagerank",
"author": "Nouamane Arhachoui",
"author_email": "nouamane.arhachoui@lip6.fr",
"download_url": "https://files.pythonhosted.org/packages/19/e2/7487fc0f2428f9daa114864682d78736bc70304a0061dd3f4d960a4499ef/psi_score-1.0.0.tar.gz",
"platform": null,
"description": "# Psi-score\n\n$\\psi$-score: Metric of user influence in Online Social Networks\n\n## Requirements\n* Python >=3.9,<3.11\n\n## Installation\n\n```bash\n$ pip install psi-score\n```\n\n## Usage\n\n```python\n>>> from psi_score import PsiScore\n>>> adjacency = {0: [1, 3], 1: [0, 2], 2: [0, 1, 3], 3: [0]}\n>>> lambdas = [0.23, 0.50, 0.86, 0.19]\n>>> mus = [0.42, 0.17, 0.10, 0.37]\n>>> psiscore = PsiScore()\n>>> scores = psiscore.fit_transform(adjacency, lambdas, mus)\n>>> scores\narray([0.21158803, 0.35253745, 0.28798439, 0.14789014])\n>>> np.round(scores, 2)\narray([0.21, 0.35, 0.29, 0.15])\n```\nYou can use another algorithm and change some parameters:\n```python\n>>> psiscore = PsiScore(solver='power_nf', n_iter=500, tol=1e-3)\n>>> scores = psiscore.fit_transform(adjacency, lambdas, mus, ps=[1], qs=[0, 3])\n```\nThe ``ps`` and ``qs`` parameters allows to have some chosen ``p_i`` and ``q_i`` vectors (only with the ``push`` and ``power_nf`` methods):\n```python\n>>> psiscore.P\n{1: array([0.5333334 , 0.1681094 , 0.46801851, 0.34442264])}\n>>> psiscore.Q\n{0: array([0.46164044, 0.0514935 , 0.02798624, 0.30484491]),\n 3: array([0.13087053, 0.01616898, 0.01850541, 0.42554885])}\n```\n\n## License\n\n`psi-score` was created by Nouamane Arhachoui. It is licensed under the terms of the MIT license.\n\n## References\n\n* Giovanidis, A., Baynat, B., Magnien, C., & Vendeville, A. (2021).\n Ranking Online Social Users by Their Influence. \n IEEE/ACM Transactions on Networking, 29(5), 2198-2214. https://doi.org/10.1109/tnet.2021.3085201\n\n* Arhachoui, N., Bautista, E., Danisch, M., & Giovanidis, A. (2022). \n A Fast Algorithm for Ranking Users by their Influence in Online Social Platforms. \n 2022 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM), 526-533. \n https://doi.org/10.1109/ASONAM55673.2022.10068673\n",
"bugtrack_url": null,
"license": "MIT",
"summary": "Metric of user influence in Online Social Networks",
"version": "1.0.0",
"split_keywords": [
"graph algorithms",
"network science",
"markov chain",
"social networks",
"pagerank"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "3e7c4663cde288d542e3bc4ac50efbed3a84038eb1d1489ca0cd2944ab6b340b",
"md5": "b1f06ea7a0de1ca996cc4ab77f7247bb",
"sha256": "8976f6bf915b2f4d429a7f977dff45a7bc4ba19c6a64941598ea8e5c00a703eb"
},
"downloads": -1,
"filename": "psi_score-1.0.0-py3-none-any.whl",
"has_sig": false,
"md5_digest": "b1f06ea7a0de1ca996cc4ab77f7247bb",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.9,<3.11",
"size": 12785,
"upload_time": "2023-04-18T15:16:58",
"upload_time_iso_8601": "2023-04-18T15:16:58.069091Z",
"url": "https://files.pythonhosted.org/packages/3e/7c/4663cde288d542e3bc4ac50efbed3a84038eb1d1489ca0cd2944ab6b340b/psi_score-1.0.0-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "19e27487fc0f2428f9daa114864682d78736bc70304a0061dd3f4d960a4499ef",
"md5": "9d3474ba0be5f31c8975344ce2e87e05",
"sha256": "c8991b213ad2d3e4537e28c20097d04eecfd1e5adc41e807fa9d3556e1c440be"
},
"downloads": -1,
"filename": "psi_score-1.0.0.tar.gz",
"has_sig": false,
"md5_digest": "9d3474ba0be5f31c8975344ce2e87e05",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.9,<3.11",
"size": 8302,
"upload_time": "2023-04-18T15:16:59",
"upload_time_iso_8601": "2023-04-18T15:16:59.816506Z",
"url": "https://files.pythonhosted.org/packages/19/e2/7487fc0f2428f9daa114864682d78736bc70304a0061dd3f4d960a4499ef/psi_score-1.0.0.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2023-04-18 15:16:59",
"github": true,
"gitlab": false,
"bitbucket": false,
"github_user": "NouamaneA",
"github_project": "psi-score",
"travis_ci": false,
"coveralls": false,
"github_actions": false,
"lcname": "psi-score"
}
JSON
