broaden

broaden(key, x, percentage=0.1)

Broaden data using gaussian noise by a specified percentage.

The broadening is relative to the standard deviation of each column in the data. Since standard deviations add with the square, random noise is added with noise_std = data_std * (sqrt ((1.0 + percentage)**2) - 1.0).

Parameters:

• x (jnp.ndarray) – Input data.

• percentage (float) – The percentage to broaden the data by.

Returns:

The broadened data.

Return type:

jnp.ndarray