rolling window sequence

path: mlstars.custom.timeseries_preprocessing.rolling_window_sequences

description: this primitive generates many sub-sequences of the original sequence. it uses a rolling window approach to create the sub-sequences out of time series data.

see json.

argument	type	description
parameters
X	numpy.ndarray	n-dimensional sequence to iterate over
index	numpy.ndarray	array containing the index values of X
drop	numpy.ndarray, str, float, bool, or None	array of boolean values indicating which value should be dropped
hyperparameters
window_size	int	length of the input sequences
target_size	int	length of the target sequences
step_size	int	indicating the number of steps to move the window forward each round
target_column	int	indicating which column of X is the target
drop_windows	bool	indicates whether the dropping functionality should be enabled
output
X	numpy.ndarray	input sequences
y	numpy.ndarray	target sequences
index	numpy.ndarray	first index value of each input sequence
target_index	numpy.ndarray	first index value of each target sequence

In [1]: import numpy as np

In [2]: from mlstars import load_primitive

In [3]: primitive = load_primitive('mlstars.custom.timeseries_preprocessing.rolling_window_sequences',
   ...:     arguments={"window_size": 10, "target_size": 1, "step_size": 1, "target_column": 0})
   ...: 

In [4]: X = np.array([1] * 50).reshape(-1, 1)

In [5]: index = np.array(range(50)).reshape(-1, 1)

In [6]: X, y, index, target_index = primitive.produce(X=X, index=index)

In [7]: print("X shape = {}\ny shape = {}\nindex shape = {}\ntarget index shape = {}".format(
   ...:     X.shape, y.shape, index.shape, target_index.shape))
   ...: 
X shape = (40, 10, 1)
y shape = (40, 1)
index shape = (40, 1)
target index shape = (40, 1)

intervals to mask fillna