The input-output dataset

The IODataset proposes two particular group names, namely INPUT_GROUP and OUTPUT_GROUP. This particular Dataset is useful for supervised machine learning and sensitivity analysis.

from __future__ import annotations

from gemseo.datasets.io_dataset import IODataset

First, we instantiate the IODataset:

dataset = IODataset()

and add some input and output variables using the methods add_input_variable() and add_output_variable() that are based on Dataset.add_variable():

dataset.add_input_variable("a", [[1.0, 2.0], [4.0, 5.0]])
dataset.add_input_variable("b", [[3.0], [6.0]])
dataset.add_output_variable("c", [[-1.0], [-2.0]])

as well as another variable:

dataset.add_variable("x", [[10.0], [20.0]])
dataset

GROUP	inputs			outputs	parameters
VARIABLE	a		b	c	x
COMPONENT	0	1	0	0	0
0	1.0	2.0	3.0	-1.0	10.0
1	4.0	5.0	6.0	-2.0	20.0

We could also do the same with the methods add_input_group() and add_output_group().

dataset = IODataset()
dataset.add_input_group(
    [[1.0, 2.0, 3.0], [4.0, 5.0, 6.0]], ["a", "b"], {"a": 2, "b": 1}
)
dataset.add_output_group([[-1.0], [-2.0]], ["c"])
dataset.add_variable("x", [[10.0], [20.0]])
dataset

GROUP	inputs			outputs	parameters
VARIABLE	a		b	c	x
COMPONENT	0	1	0	0	0
0	1.0	2.0	3.0	-1.0	10.0
1	4.0	5.0	6.0	-2.0	20.0

Then, we can easily access the names of the input and output variables:

dataset.input_names, dataset.output_names

(['a', 'b'], ['c'])

and those of all variables:

dataset.variable_names

['a', 'b', 'c', 'x']

The IODataset provides also the number of samples:

dataset.n_samples

2

and the samples:

dataset.samples

[0, 1]

Lastly, we can get the input data as an IODataset view:

dataset.input_dataset

GROUP	inputs
VARIABLE	a		b
COMPONENT	0	1	0
0	1.0	2.0	3.0
1	4.0	5.0	6.0

and the same for the output data:

dataset.output_dataset

GROUP	outputs
VARIABLE	c
COMPONENT	0
0	-1.0
1	-2.0