sf.apps.data.QM9Exact

class QM9Exact[source]

Bases: strawberryfields.apps.data.feature.FeatureDataset

Exactly-calculated feature vectors of 1100 randomly-chosen molecules from the QM9 dataset.

The QM9 dataset is widely used in benchmarking performance of machine learning models in estimating molecular properties [6], [7].

Coulomb matrices were used as adjacency matrices to represent molecules in this case.

The Monte-Carlo estimated feature vectors of certain events of these 1100 molecules are also available in the QM9MC class.

method = "exact"
n_mean = 6
unit = "orbits"
unit_data = [[1, 1], [2], [1, 1, 1, 1], [2, 1, 1], [2, 2], [1, 1, 1, 1, 1, 1], [2, 1, 1, 1, 1], [2, 2, 1, 1], [2, 2, 2]]

method

n_mean

unit

unit_data

method = 'exact'
n_mean = 6
unit = 'orbits'
unit_data = [[1, 1], [2], [1, 1, 1, 1], [2, 1, 1], [2, 2], [1, 1, 1, 1, 1, 1], [2, 1, 1, 1, 1], [2, 2, 1, 1], [2, 2, 2]]