# Utils

## Symbolization

The following methods are used by this library to compute entropies.

`Entropies.GaussianSymbolization`

— Type`GaussianSymbolization(; n_categories::Int = 5)`

A symbolization scheme where the elements of `x`

, which is the time series $x = \{x_j, j = 1, 2, \ldots, N \}$, is to a new time series $y = \{y_j, j = 1, 2, \ldots, N \}$ where $y_j \in (0, 1)$ and

\[y_j = \dfrac{1}{\sigma\sqrt{2\pi}} \int_{x = -\infty}^{x_j} \exp{\dfrac{-{(x-\mu)}^2}{2{\sigma}^2}} dx,\]

where $\mu$ and $\sigma$ are the mean and standard deviations of `x`

.

See also: `Entropies.symbolize`

.