# colormaps { #tvbo.adapters.colormaps }

`adapters.colormaps`

Colormaps a figure can name that the plotting stack does not ship.

A replication reproduces a published figure's colour convention, and the convention is usually whatever the authors' tool defaulted to. Where that map exists in matplotlib the spec just names it; where it does not, every study otherwise carries its own copy of the same anchor table — so the tables live here once and :func:`resolve` turns any declared name into something the backend accepts.

Adding a map here is a format-support decision, not a styling one: it makes a NAME resolvable, and which map a figure uses stays entirely in the spec.

``parula`` is MATLAB's default since R2014b, and so the colour convention of a large share of published neuroimaging figures. Neither matplotlib nor bsplot ships it, and bsplot's ``fake_parula`` is a visibly different map — it starts violet (0.267, 0.033, 0.619) where parula starts blue — so a figure reproduced with the lookalike is off from its original at every value.

## Attributes

| Name | Description |
| --- | --- |
| [PARULA](#tvbo.adapters.colormaps.PARULA) |  |
| [TABLES](#tvbo.adapters.colormaps.TABLES) |  |

## Functions

| Name | Description |
| --- | --- |
| [resolve](#tvbo.adapters.colormaps.resolve) | A declared colormap name as something the plotting backend accepts (public API). |

### resolve { #tvbo.adapters.colormaps.resolve }

```python
adapters.colormaps.resolve(name)
```

A declared colormap name as something the plotting backend accepts (public API).

A name the backend already knows is passed through untouched — matplotlib's and bsplot's own catalogues stay authoritative — so this only fills the gaps in :data:`TABLES`. ``None`` passes through as ``None`` (the backend's default).