Metadata-Version: 2.1
Name: savvi
Version: 0.2.0
Summary: savvi is a package for Safe Anytime Valid Inference
Author-Email: =?utf-8?b?THXDrXMgQXNzdW7Dp8Ojbw==?= <assuncaolfi@gmail.com>
License: MIT
Requires-Python: >=3.11
Requires-Dist: numpy<2.0.0
Requires-Dist: scipy>=1.13.1
Requires-Dist: cvxpy>=1.5.2
Description-Content-Type: text/markdown

# savvi


<!-- badges: start -->
<!-- badges: end -->

savvi is a package for Safe Anytime Valid Inference. Also, it’s a savvy
pun.

From Ramdas et al. ([2023](#ref-ramdas2023gametheoretic)):

> Safe anytime-valid inference (SAVI) provides measures of statistical
> evidence and certainty – e-processes for testing and confidence
> sequences for estimation – that remain valid at all stopping times,
> accommodating continuous monitoring and analysis of accumulating data
> and optional stopping or continuation for any reason.

## Install

Stable version:

``` shell
pip install savvi
```

Development version:

``` shell
pip install git+https://github.com/assuncaolfi/savvi
```

For development, use [pdm](https://github.com/pdm-project/pdm).

## Get started

See the [example
gallery](https://assuncaolfi.github.io/savvi/examples/).

<div id="refs" class="references csl-bib-body hanging-indent"
entry-spacing="0">

<div id="ref-ramdas2023gametheoretic" class="csl-entry">

Ramdas, Aaditya, Peter Grünwald, Vladimir Vovk, and Glenn Shafer. 2023.
“Game-Theoretic Statistics and Safe Anytime-Valid Inference.”
<https://arxiv.org/abs/2210.01948>.

</div>

</div>
