make_arima

cuml.datasets.make_arima(

batch_size=1000,

n_obs=100,

order=(1, 1, 1),

seasonal_order=(0, 0, 0, 0),

intercept=False,

random_state=None,

dtype='float64',

)

Generates a dataset of time series by simulating an ARIMA process of a given order.

Deprecated since version 26.08: cuml.datasets.make_arima is deprecated and will be removed in the cuML 26.12 release.

Parameters:

batch_size: int

Number of time series to generate

n_obs: int

Number of observations per series

orderTuple[int, int, int]

Order (p, d, q) of the simulated ARIMA process

seasonal_order: Tuple[int, int, int, int]

Seasonal ARIMA order (P, D, Q, s) of the simulated ARIMA process

intercept: bool or int

Whether to include a constant trend mu in the simulated ARIMA process

random_state: int, RandomState instance or None (default)

Seed for the random number generator for dataset creation.

dtype: string or numpy dtype (default: ‘float64’)

The output dtype. Only float32 or float64 supported.

Returns:

out: array-like, shape (n_obs, batch_size)

Array of the requested type containing the generated dataset

Examples

from cuml.datasets import make_arima
y = make_arima(1000, 100, (2,1,2), (0,1,2,12), 0)