# TabPFN Common Utilities
A comprehensive utility package for [TabPFN](https://github.com/priorlabs/tabpfn) - the foundation model for tabular data.
## Features
### 🔒 Privacy-First Telemetry System
- **Anonymous & Aggregated Data Collection**: Implements safe, GDPR-compliant telemetry that respects user privacy
- **Configurable Analytics**: Optional telemetry that can be disabled via environment variables
- **Usage Pattern Insights**: Tracks TabPFN usage patterns to improve the model and user experience
- **Zero Personal Data**: No personal information or sensitive data is collected or transmitted
### 💰 Cost Estimation
- **Resource Planning**: Accurate estimation of computational costs and duration for TabPFN predictions
- **Cloud Pricing**: Essential for resource planning in cloud-based TabPFN services
- **Task-Specific Calculations**: Different cost models for classification vs regression tasks
### 📊 Data Processing Utilities
- **Regression Results**: Comprehensive handling of prediction outputs with mean, median, mode, and quantiles
- **Data Serialization**: Convert between pandas DataFrames, NumPy arrays, and CSV formats
- **Dataset Management**: Load and preprocess standard ML datasets with proper train/test splits
- **Preprocessing Configuration**: Extensive options for data transformation strategies
## Installation
```bash
pip install tabpfn-common-utils
```
Or with uv:
```bash
uv add tabpfn-common-utils
```
## Quick Start
### Telemetry (Privacy-Compliant)
```python
from tabpfn_common_utils.telemetry import ProductTelemetry
# Initialize telemetry service (anonymous, GDPR-compliant)
telemetry = ProductTelemetry()
# Track usage events (no personal data collected)
telemetry.capture(...)
# Telemetry can be disabled by setting environment variable
export TABPFN_DISABLE_TELEMETRY=1
```
### Regression Results
```python
from tabpfn_common_utils.regression_pred_result import RegressionPredictResult
# Handle regression prediction results
result = RegressionPredictResult({
"mean": [1.2, 2.3, 3.4],
"median": [1.1, 2.2, 3.3],
"mode": [1.0, 2.0, 3.0],
"quantile_0.25": [0.9, 1.9, 2.9],
"quantile_0.75": [1.5, 2.5, 3.5]
})
# Convert to basic representation for serialization
basic_repr = RegressionPredictResult.to_basic_representation(result)
```
### Data Utilities
```python
from tabpfn_common_utils.utils import get_example_dataset, serialize_to_csv_formatted_bytes
import pandas as pd
# Load example dataset
X_train, X_test, y_train, y_test = get_example_dataset("iris")
# Serialize data to CSV bytes
csv_bytes = serialize_to_csv_formatted_bytes(X_train)
```
## Privacy & Compliance
This package implements **privacy-first telemetry** that:
- ✅ **GDPR Compliant**: No personal data collection
- ✅ **Anonymous Only**: No user identification or tracking
- ✅ **Aggregated Data**: Only statistical insights are collected
- ✅ **User Control**: Can be completely disabled
- ✅ **Transparent**: Open source code for full transparency
Telemetry data helps improve TabPFN but never compromises user privacy.
## Development
### Setup
```bash
# Install dependencies
uv sync
# Activate virtual environment
source .venv/bin/activate
# Run tests
uv run pytest
# Type checking
uv run pyright
# Code formatting
uv run ruff check --fix
```
### Adding Dependencies
```bash
# Add runtime dependency
uv add <package_name>
# Add development dependency
uv add --group dev <package_name>
```
## Contributing
Contributions are welcome! Please ensure all code passes type checking and formatting requirements.
## Links
- [TabPFN Main Repository](https://github.com/priorlabs/tabpfn)
- [Documentation](https://github.com/priorlabs/tabpfn_common_utils)
- [Issues](https://github.com/priorlabs/tabpfn_common_utils/issues)
Raw data
{
"_id": null,
"home_page": null,
"name": "tabpfn-common-utils",
"maintainer": null,
"docs_url": null,
"requires_python": ">=3.9",
"maintainer_email": "Prior Labs <hello@priorlabs.ai>",
"keywords": "machine-learning, tabpfn, utilities",
"author": null,
"author_email": "Prior Labs <hello@priorlabs.ai>",
"download_url": "https://files.pythonhosted.org/packages/44/88/38697306023717cbca0a25cdfdff63c88cefa23519dbc5a343265b7fd95c/tabpfn_common_utils-0.1.5.tar.gz",
"platform": null,
"description": "# TabPFN Common Utilities\n\nA comprehensive utility package for [TabPFN](https://github.com/priorlabs/tabpfn) - the foundation model for tabular data.\n\n## Features\n\n### \ud83d\udd12 Privacy-First Telemetry System\n- **Anonymous & Aggregated Data Collection**: Implements safe, GDPR-compliant telemetry that respects user privacy\n- **Configurable Analytics**: Optional telemetry that can be disabled via environment variables\n- **Usage Pattern Insights**: Tracks TabPFN usage patterns to improve the model and user experience\n- **Zero Personal Data**: No personal information or sensitive data is collected or transmitted\n\n### \ud83d\udcb0 Cost Estimation\n- **Resource Planning**: Accurate estimation of computational costs and duration for TabPFN predictions\n- **Cloud Pricing**: Essential for resource planning in cloud-based TabPFN services\n- **Task-Specific Calculations**: Different cost models for classification vs regression tasks\n\n### \ud83d\udcca Data Processing Utilities\n- **Regression Results**: Comprehensive handling of prediction outputs with mean, median, mode, and quantiles\n- **Data Serialization**: Convert between pandas DataFrames, NumPy arrays, and CSV formats\n- **Dataset Management**: Load and preprocess standard ML datasets with proper train/test splits\n- **Preprocessing Configuration**: Extensive options for data transformation strategies\n\n## Installation\n\n```bash\npip install tabpfn-common-utils\n```\n\nOr with uv:\n```bash\nuv add tabpfn-common-utils\n```\n\n## Quick Start\n\n### Telemetry (Privacy-Compliant)\n\n```python\nfrom tabpfn_common_utils.telemetry import ProductTelemetry\n\n# Initialize telemetry service (anonymous, GDPR-compliant)\ntelemetry = ProductTelemetry()\n\n# Track usage events (no personal data collected)\ntelemetry.capture(...)\n\n# Telemetry can be disabled by setting environment variable\nexport TABPFN_DISABLE_TELEMETRY=1\n```\n\n### Regression Results\n\n```python\nfrom tabpfn_common_utils.regression_pred_result import RegressionPredictResult\n\n# Handle regression prediction results\nresult = RegressionPredictResult({\n \"mean\": [1.2, 2.3, 3.4],\n \"median\": [1.1, 2.2, 3.3],\n \"mode\": [1.0, 2.0, 3.0],\n \"quantile_0.25\": [0.9, 1.9, 2.9],\n \"quantile_0.75\": [1.5, 2.5, 3.5]\n})\n\n# Convert to basic representation for serialization\nbasic_repr = RegressionPredictResult.to_basic_representation(result)\n```\n\n### Data Utilities\n\n```python\nfrom tabpfn_common_utils.utils import get_example_dataset, serialize_to_csv_formatted_bytes\nimport pandas as pd\n\n# Load example dataset\nX_train, X_test, y_train, y_test = get_example_dataset(\"iris\")\n\n# Serialize data to CSV bytes\ncsv_bytes = serialize_to_csv_formatted_bytes(X_train)\n```\n\n## Privacy & Compliance\n\nThis package implements **privacy-first telemetry** that:\n\n- \u2705 **GDPR Compliant**: No personal data collection\n- \u2705 **Anonymous Only**: No user identification or tracking\n- \u2705 **Aggregated Data**: Only statistical insights are collected\n- \u2705 **User Control**: Can be completely disabled\n- \u2705 **Transparent**: Open source code for full transparency\n\nTelemetry data helps improve TabPFN but never compromises user privacy.\n\n## Development\n\n### Setup\n\n```bash\n# Install dependencies\nuv sync\n\n# Activate virtual environment\nsource .venv/bin/activate\n\n# Run tests\nuv run pytest\n\n# Type checking\nuv run pyright\n\n# Code formatting\nuv run ruff check --fix\n```\n\n### Adding Dependencies\n\n```bash\n# Add runtime dependency\nuv add <package_name>\n\n# Add development dependency\nuv add --group dev <package_name>\n```\n\n## Contributing\n\nContributions are welcome! Please ensure all code passes type checking and formatting requirements.\n\n## Links\n\n- [TabPFN Main Repository](https://github.com/priorlabs/tabpfn)\n- [Documentation](https://github.com/priorlabs/tabpfn_common_utils)\n- [Issues](https://github.com/priorlabs/tabpfn_common_utils/issues)",
"bugtrack_url": null,
"license": null,
"summary": "Utilities shared between TabPFN codebases",
"version": "0.1.5",
"project_urls": {
"Homepage": "https://github.com/priorlabs/tabpfn_common_utils",
"Issues": "https://github.com/priorlabs/tabpfn_common_utils/issues"
},
"split_keywords": [
"machine-learning",
" tabpfn",
" utilities"
],
"urls": [
{
"comment_text": null,
"digests": {
"blake2b_256": "5f591b29905f0b814e791d55486abed71cd399b9aceddc1ce504ee90610e7c22",
"md5": "abc2a6b2e5c3be3ed1b840a0da3567a9",
"sha256": "a56b5eea2388e7f8fe9bece365e54e73cfaeceb9b4c9edbdde57b4947d41f847"
},
"downloads": -1,
"filename": "tabpfn_common_utils-0.1.5-py3-none-any.whl",
"has_sig": false,
"md5_digest": "abc2a6b2e5c3be3ed1b840a0da3567a9",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.9",
"size": 10419,
"upload_time": "2025-09-05T09:49:21",
"upload_time_iso_8601": "2025-09-05T09:49:21.680043Z",
"url": "https://files.pythonhosted.org/packages/5f/59/1b29905f0b814e791d55486abed71cd399b9aceddc1ce504ee90610e7c22/tabpfn_common_utils-0.1.5-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": null,
"digests": {
"blake2b_256": "448838697306023717cbca0a25cdfdff63c88cefa23519dbc5a343265b7fd95c",
"md5": "18af18539d44fefc22e950c8d6fb99cd",
"sha256": "8b762753455812cd2eb938f46f0fdbdc7b2f69feaa3b4b4bf28888edea38ec29"
},
"downloads": -1,
"filename": "tabpfn_common_utils-0.1.5.tar.gz",
"has_sig": false,
"md5_digest": "18af18539d44fefc22e950c8d6fb99cd",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.9",
"size": 1908199,
"upload_time": "2025-09-05T09:49:26",
"upload_time_iso_8601": "2025-09-05T09:49:26.378448Z",
"url": "https://files.pythonhosted.org/packages/44/88/38697306023717cbca0a25cdfdff63c88cefa23519dbc5a343265b7fd95c/tabpfn_common_utils-0.1.5.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2025-09-05 09:49:26",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "priorlabs",
"github_project": "tabpfn_common_utils",
"travis_ci": false,
"coveralls": false,
"github_actions": true,
"lcname": "tabpfn-common-utils"
}
JSON
