# Installation

```bash
pip install numeraire-dataset
```

This installs the tidyfinance-backed frame loaders and the self-built builders. Python 3.11+ is
required; [tidyfinance](https://github.com/tidy-finance/py-tidyfinance) is a hard dependency (the
primary backend for standard sources).

## Extras

```bash
pip install "numeraire-dataset[numeraire]"   # + the point-in-time view helpers (numeraire bridge)
pip install "numeraire-dataset[wrds]"        # + live CRSP / Compustat pulls (your own WRDS account)
```

The plain frame loaders (`load_ff_factors`, `load_goyal_welch`) carry **no numeraire dependency** and
are usable on their own. The `[numeraire]` extra adds the `*_view` helpers that return a numeraire
{class}`~numeraire.core.data.TimeSeriesView`. The framework can also pull this package in for you:

```bash
pip install "numeraire[data]"   # numeraire + numeraire-dataset
```

With [uv](https://docs.astral.sh/uv/):

```bash
uv add numeraire-dataset
```

## Verify

```python
from importlib.metadata import version

import numeraire_dataset  # noqa: F401

print(version("numeraire-dataset"))
```