USPTO

Yields

Dataset Description

TDC parses the yields outcome from the full USPTO (United States Patent and Trademark Office) dataset.

Task Description

Given reactant and product set X, predict the yields Y.

Dataset Statistics

853,638 reactions.

Available Splits

Random Split

Usage Example

from tdc_ml.single_pred import Yields

data = Yields(name='USPTO')

# Access the data
df = data.get_data()
print(df.head())

# Get train/val/test splits
split = data.get_split()
print(split)

License

This dataset is licensed under CC0