# bdi-viz
[](https://github.com/VIDA-NYU/bdi-viz/actions/workflows/build.yml)
[](https://github.com/VIDA-NYU/bdi-viz/actions/workflows/lint.yml)
[](https://bdi-viz.readthedocs.io/en/latest/)
## Contents
- [1. Introduction](#sparkle-1-introduction)
- [2. Installation](#package-2-installation)
- [3. Quick Start](#rocket-3-quick-start)
- [4. Documentation](#page_facing_up-4-documentation)
- [4.1 Read the Docs](#41-read-the-docs)
- [4.2 Demo Video](#42-demo-video)
## :sparkle: 1. Introduction
BDIViz is a powerful, interactive tool designed as an extension to [BDIKit](https://github.com/VIDA-NYU/bdi-kit) to assist biomedical researchers and domain experts in performing schema matching tasks. Built to address the challenges of matching complex biomedical datasets, BDIViz leverages a visual approach to streamline the process and enhance both speed and accuracy.
Key features of BDIViz include:
- **Interactive Heatmap** for exploring and comparing matching candidates.
- **Value Comparisons** Panel for analyzing similarities between attributes.
- **Detailed Analysis** Panel offering in-depth insights into attribute value distributions.
- **Filtering & Refinement Tools** to customize and adjust matching candidates based on datatype and similarity scores.
- **Expert-in-the-Loop Workflow** allowing users to iteratively accept, reject, or refine matches, keeping the expert in control of decision-making.
BDIViz is designed to be integrated with Python notebooks, providing a flexible and easy-to-use tool for domain-specific schema matching in biomedical research and beyond.
## :package: 2. Installation
To use ``BDI-Viz``, install it using pip:
```bash
pip install bdi-viz
```
## :rocket: 3. Quick Start
``BDI-Viz 1.0`` is built leveraging [Panel](https://panel.holoviz.org/). The application is designed to provide a user-friendly interface on jupyter notebooks. Where users can explore the schema matching recommandations, interact with the result, and pass them to the next step of the data integration process.
```python
import pandas as pd
from bdiviz import BDISchemaMatchingHeatMap
# Load the data
source_df = pd.read_csv('data/source.csv')
target_df = pd.read_csv('data/target.csv')
# Render the BDI-Viz Heatmap
heatmap_manager = BDISchemaMatchingHeatMap(
source=source_df,
target=target_df,
top_k=20,
)
heatmap_manager.plot_heatmap()
```
The following interface will be displayed in the jupyter notebook:
## :page_facing_up: 4. Documentation
### 4.1 Read the Docs
For more information, please refer to the [documentation](https://bdi-viz.readthedocs.io/en/latest/).
### 4.2 Demo Video
[BDIViz Demo](https://drive.google.com/file/d/1eAbDicO0oXIbbVg56m3H8xdNDDsBGBLI/view?usp=drive_link)
Raw data
{
"_id": null,
"home_page": "https://github.com/VIDA-NYU/bdi-viz",
"name": "bdi-viz",
"maintainer": "Eden Wu",
"docs_url": null,
"requires_python": null,
"maintainer_email": "eden.wu@nyu.edu",
"keywords": "askem, data integration, nyu",
"author": "Eden Wu",
"author_email": "eden.wu@nyu.edu",
"download_url": "https://files.pythonhosted.org/packages/a6/29/de80ed098ac225467734666eb8c3f87014f44df2b269420ffc526122bb82/bdi_viz-0.1.0.dev0.tar.gz",
"platform": null,
"description": "# bdi-viz\n[](https://github.com/VIDA-NYU/bdi-viz/actions/workflows/build.yml)\n[](https://github.com/VIDA-NYU/bdi-viz/actions/workflows/lint.yml)\n[](https://bdi-viz.readthedocs.io/en/latest/)\n\n\n## Contents\n\n- [1. Introduction](#sparkle-1-introduction)\n- [2. Installation](#package-2-installation)\n- [3. Quick Start](#rocket-3-quick-start)\n- [4. Documentation](#page_facing_up-4-documentation)\n - [4.1 Read the Docs](#41-read-the-docs)\n - [4.2 Demo Video](#42-demo-video)\n\n## :sparkle: 1. Introduction\n\nBDIViz is a powerful, interactive tool designed as an extension to [BDIKit](https://github.com/VIDA-NYU/bdi-kit) to assist biomedical researchers and domain experts in performing schema matching tasks. Built to address the challenges of matching complex biomedical datasets, BDIViz leverages a visual approach to streamline the process and enhance both speed and accuracy.\n\nKey features of BDIViz include:\n\n- **Interactive Heatmap** for exploring and comparing matching candidates.\n- **Value Comparisons** Panel for analyzing similarities between attributes.\n- **Detailed Analysis** Panel offering in-depth insights into attribute value distributions.\n- **Filtering & Refinement Tools** to customize and adjust matching candidates based on datatype and similarity scores.\n- **Expert-in-the-Loop Workflow** allowing users to iteratively accept, reject, or refine matches, keeping the expert in control of decision-making.\n\nBDIViz is designed to be integrated with Python notebooks, providing a flexible and easy-to-use tool for domain-specific schema matching in biomedical research and beyond.\n\n## :package: 2. Installation\n\nTo use ``BDI-Viz``, install it using pip:\n\n```bash\npip install bdi-viz\n```\n\n\n## :rocket: 3. Quick Start\n``BDI-Viz 1.0`` is built leveraging [Panel](https://panel.holoviz.org/). The application is designed to provide a user-friendly interface on jupyter notebooks. Where users can explore the schema matching recommandations, interact with the result, and pass them to the next step of the data integration process.\n\n```python\nimport pandas as pd\nfrom bdiviz import BDISchemaMatchingHeatMap\n\n# Load the data\nsource_df = pd.read_csv('data/source.csv')\ntarget_df = pd.read_csv('data/target.csv')\n\n# Render the BDI-Viz Heatmap\nheatmap_manager = BDISchemaMatchingHeatMap(\n source=source_df,\n target=target_df,\n top_k=20,\n)\nheatmap_manager.plot_heatmap()\n```\n\nThe following interface will be displayed in the jupyter notebook:\n\n\n\n## :page_facing_up: 4. Documentation\n\n### 4.1 Read the Docs\nFor more information, please refer to the [documentation](https://bdi-viz.readthedocs.io/en/latest/).\n\n### 4.2 Demo Video\n[BDIViz Demo](https://drive.google.com/file/d/1eAbDicO0oXIbbVg56m3H8xdNDDsBGBLI/view?usp=drive_link)\n\n",
"bugtrack_url": null,
"license": "Apache-2.0",
"summary": "bdi-viz library",
"version": "0.1.0.dev0",
"project_urls": {
"Homepage": "https://github.com/VIDA-NYU/bdi-viz"
},
"split_keywords": [
"askem",
" data integration",
" nyu"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "e9a034d790cbbe8bb0f77ef1839637989f5ec87700fe20ca339042eb13f4b36b",
"md5": "dc5cd68d950997af2b167fcec7b84c7b",
"sha256": "90f67f3addd222132a62aa883411e7be140bb0e9781f3ab3f1daee8ddf894553"
},
"downloads": -1,
"filename": "bdi_viz-0.1.0.dev0-py3-none-any.whl",
"has_sig": false,
"md5_digest": "dc5cd68d950997af2b167fcec7b84c7b",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": null,
"size": 18515,
"upload_time": "2025-01-07T01:38:33",
"upload_time_iso_8601": "2025-01-07T01:38:33.285502Z",
"url": "https://files.pythonhosted.org/packages/e9/a0/34d790cbbe8bb0f77ef1839637989f5ec87700fe20ca339042eb13f4b36b/bdi_viz-0.1.0.dev0-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "a629de80ed098ac225467734666eb8c3f87014f44df2b269420ffc526122bb82",
"md5": "1fd3532bce74fe75137772ee9cd23411",
"sha256": "5e145e476919b7aea4bf0dcce55b0fdf09da8b889f9fc17160f2a3a61490a3e3"
},
"downloads": -1,
"filename": "bdi_viz-0.1.0.dev0.tar.gz",
"has_sig": false,
"md5_digest": "1fd3532bce74fe75137772ee9cd23411",
"packagetype": "sdist",
"python_version": "source",
"requires_python": null,
"size": 18621,
"upload_time": "2025-01-07T01:38:34",
"upload_time_iso_8601": "2025-01-07T01:38:34.408136Z",
"url": "https://files.pythonhosted.org/packages/a6/29/de80ed098ac225467734666eb8c3f87014f44df2b269420ffc526122bb82/bdi_viz-0.1.0.dev0.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2025-01-07 01:38:34",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "VIDA-NYU",
"github_project": "bdi-viz",
"travis_ci": false,
"coveralls": false,
"github_actions": true,
"requirements": [
{
"name": "altair",
"specs": []
},
{
"name": "bdi-kit",
"specs": []
},
{
"name": "bokeh",
"specs": []
},
{
"name": "datamart_profiler",
"specs": []
},
{
"name": "panel",
"specs": [
[
"!=",
"1.4.3"
]
]
}
],
"lcname": "bdi-viz"
}
JSON
