https://github.com/vcerqueira/cardtale
Data Cards for Time Series
Science Score: 13.0%
This score indicates how likely this project is to be science-related based on various indicators:
-
○CITATION.cff file
-
✓codemeta.json file
Found codemeta.json file -
○.zenodo.json file
-
○DOI references
-
○Academic publication links
-
○Committers with academic emails
-
○Institutional organization owner
-
○JOSS paper metadata
-
○Scientific vocabulary similarity
Low similarity (16.7%) to scientific vocabulary
Keywords
Repository
Data Cards for Time Series
Basic Info
Statistics
- Stars: 3
- Watchers: 2
- Forks: 1
- Open Issues: 1
- Releases: 0
Topics
Metadata Files
README.md
cardtale
cardtale is a Python package for generating automated model and data cards for time series, streamlining the documentation process for machine learning models and datasets.
Key Features
- Automated generation of PDF reports with comprehensive time series analysis
- Built-in statistical analysis and visualization of temporal patterns
- Support for univariate time series data
Each time series is studied from multiple dimensions, including: - Data Overview: Fundamental characteristics and statistical properties analysis - Trend Analysis: Long-term growth patterns and level stabilization assessment - Seasonality Detection: Analysis of multiple seasonality levels with strength metrics - Variance Analysis: Heteroskedasticity testing and variance stabilization methods - Change Point Detection: Identification of structural changes and their impact
Basic Example
Here's a basic example of cardtale.
```python from datasetsforecast.m3 import M3 from cardtale.cards.builder import CardsBuilder
df, *_ = M3.load('./assets', group='Monthly')
freq = 'ME' uid = 'M1080'
seriesdf = df.query(f'uniqueid=="{uid}"').reset_index(drop=True)
tcard = CardsBuilder(seriesdf, freq) tcard.buildcards() tcard.get_pdf(path='example.pdf')
```
Screenshots
⚠️ WARNING
cardtale is in the early stages of development. It is designed to cover datasets containing single univariate time series, focusing on forecasting tasks.
cardtale has been developed for monthly time series. So, the output for other frequencies may not be as reliable. Especially time series with complex seasonality.
If you encounter any issues, please report them in GitHub Issues
Installation
Prerequisites
Required dependencies:
arch==7.1.0
pandas==2.2.3
lightgbm==4.5.0
neuralforecast==1.7.5
mlforecast==0.13.4
statsforecast==1.7.8
datasetsforecast==0.0.8
numerize==0.12
plotnine==0.13.6
statsmodels==0.14.4
jinja2==3.1.4
ruptures==1.1.9
weasyprint==62.3
You can install cardtale using pip:
bash
pip install cardtale -U
[Optional] Installation from source
To install cardtale from source, clone the repository and run the following command:
bash
git clone https://github.com/vcerqueira/cardtale
pip install -e cardtale
License
Cardtale is released under the MIT License. See the LICENSE file for more details.
Mission
In the machine learning lifecycle, proper documentation of models and datasets is crucial for transparency, reproducibility, and responsible AI practices. However, creating comprehensive model and data cards can be time-consuming. The Python package cardtale aims to partially automate this process, making it more efficient and consistent.
The goal of cardtale is to generate a set of analyses, visualizations, and interpretations based on input model metrics and dataset characteristics. While it doesn't replace the expertise of data scientists or domain experts, Cardtale speeds up the creation of model and data cards, guiding analysts towards key insights and areas that may require further exploration.
Project Funded by
Agenda “Center for Responsible AI”, nr. C645008882-00000055, investment project nr. 62, financed by the Recovery and Resilience Plan (PRR) and by European Union - NextGeneration EU.
Owner
- Name: Vitor Cerqueira
- Login: vcerqueira
- Kind: user
- Company: -
- Repositories: 3
- Profile: https://github.com/vcerqueira
Researcher at Dalhousie University
GitHub Events
Total
- Issues event: 1
- Watch event: 2
- Issue comment event: 3
- Push event: 83
- Fork event: 1
Last Year
- Issues event: 1
- Watch event: 2
- Issue comment event: 3
- Push event: 83
- Fork event: 1
Committers
Last synced: 6 months ago
Top Committers
| Name | Commits | |
|---|---|---|
| Vitor Cerqueira | c****l@g****m | 82 |
| Vitor Cerqueira | v****a@V****l | 5 |
| Vitor Cerqueira | v****a@v****t | 4 |
| Vitor Cerqueira | v****a@l****t | 3 |
| Vitor Cerqueira | v****a@l****t | 2 |
Committer Domains (Top 20 + Academic)
Issues and Pull Requests
Last synced: 5 months ago
All Time
- Total issues: 1
- Total pull requests: 0
- Average time to close issues: N/A
- Average time to close pull requests: N/A
- Total issue authors: 1
- Total pull request authors: 0
- Average comments per issue: 3.0
- Average comments per pull request: 0
- Merged pull requests: 0
- Bot issues: 0
- Bot pull requests: 0
Past Year
- Issues: 1
- Pull requests: 0
- Average time to close issues: N/A
- Average time to close pull requests: N/A
- Issue authors: 1
- Pull request authors: 0
- Average comments per issue: 3.0
- Average comments per pull request: 0
- Merged pull requests: 0
- Bot issues: 0
- Bot pull requests: 0
Top Authors
Issue Authors
- Sandy4321 (1)
Pull Request Authors
Top Labels
Issue Labels
Pull Request Labels
Packages
- Total packages: 1
-
Total downloads:
- pypi 36 last-month
- Total dependent packages: 0
- Total dependent repositories: 0
- Total versions: 4
- Total maintainers: 1
pypi.org: cardtale
Data, Model, and Algorithm Cards for Time Series
- Homepage: https://github.com/vcerqueira/cardtale
- Documentation: https://cardtale.readthedocs.io/
- License: MIT License
-
Latest release: 0.1.4
published 8 months ago
Rankings
Maintainers (1)
Dependencies
- mlforecast ==0.13.4
- neuralforecast ==1.7.5
- numerize ==0.12
- pandas ==2.2.3
- plotnine ==0.13.6
- rpy2 ==3.5.16
- scikit-learn ==1.5.2
- statsmodels ==0.14.4
- weasyprint ==62.3