convert labels from log-scale Args: form (str, optional): standard log-transformation or binding nM <-> p transformation.

log conversion helper Args: y (list): a list of labels Returns: np.array: an array of log-transformed labels

convert labels to log-scale Args: form (str, optional): standard log-transformation or binding nM <-> p transformation.

split within each stratification defined by the group column for training/validation split Args: train_val (pd.DataFrame): the train+valid dat

Update node representations. Parameters ---------- g : DGLGraph DGLGraph for a batch of graphs feats : Flo

Update node representations. Parameters ---------- g : DGLGraph DGLGraph for a batch of graphs node_feats

Update node representations Parameters ---------- g : DGLGraph DGLGraph for a batch of graphs categorical_

Perform the forward pass.

Args: inputs: prediction matrix (before softmax) with shape (batch_size, num_classes) targets: ground truth labels wi

Args: inputs: prediction matrix (before softmax) with shape (batch_size, num_classes) targets: ground truth labels wi

Args: inputs: prediction matrix (before softmax) with shape (batch_size, num_classes) targets: ground truth labels wi

Args: inputs: prediction matrix (before softmax) with shape (batch_size, num_classes) targets: ground truth labels wi

Args: inputs: prediction matrix (before softmax) with shape (batch_size, num_classes) targets: ground truth labels wi

Args: inputs: prediction matrix (before softmax) with shape (batch_size, num_classes) targets: ground truth labels wi

Args: inputs: prediction matrix (before softmax) with shape (batch_size, num_classes) targets: ground truth labels wi

Args: inputs: prediction matrix (before softmax) with shape (batch_size, num_classes) targets: ground truth labels wi

Args: inputs: prediction matrix (before softmax) with shape (batch_size, num_classes) targets: ground truth labels wi

Gets an activation function module given the name of the activation. :param activation: The name of the activation function. #TODO: the type

Get the args and kwargs for corresponding class. Need to be override by child class Should be written by designer and used by user

Get the config from args. :param args: the input arguments. :param params: the necessary parameters to initialize target clas

to retrieve the biomedical meaning of labels Args: name (str): the name of the dataset path (str, optional): the dataset path, where the

get the biomedical meaning of label Args: output_format (str, optional): dict/df/array for label Returns

The function to generate parser. Need to be override by child class. Should be written by designer and used by user :param pa

to retrieve the type of reactions for reaction dataset Args: name (str): dataset name path (str, optional): dataset path output_form

split dataset into train/validation/test. Args: method (str, optional): split method, the default value is 'ran

split dataset into train/validation/test. Args: method (str, optional): split method, the default value is 'rand

Arguments: method: splitting schemes, choose from random, cold_{entity}, scaffold, defaults to 'random' seed: the random seed for splitting d

Removing duplicated drug-target pairs with different binding affinities.

Initializes the weights of a model in place. :param model: An nn.Module.

plot the distribution of label Args: y (list): a list of labels name (None, optional): dataset name

visualize distribution of labels

load an object from a path Args: path (str): the path where the pickle file locates Returns: object: loaded pickle file

r""" References: Verma et al., Manifold Mixup: Better Representations by Interpolating Hidden States, ICML 2019. Speciall

r""" References: Zhang et al., mixup: Beyond Empirical Risk Minimization, ICLR

negative sampling Args: frac (int, optional): the ratio between negative and positive samples. Returns

print statistics

print the statistics of the dataset

print the statistics of the dataset

r""" Reference: Chou et al. Remix: Rebalanced Mixup, ECCV 2020 workshop. The difference between input mixup and remix is

Reinitialize model parameters.

to get all available dataset names given a task Args: name (str): the name of query task Returns: list: a list of available datasets

save an object to a pickle file Args: path (str): the path to save the pickle file obj (object): any file

Summary TODO Args: threshold (float, optional): threshold to binarize the data. format (str, optional): form

convert the results to submission-ready format in leaderboard Args: results (dict): a dictionary of metrics across five runs Returns:

Get protein sequence from Uniprot ID Args: ProteinID (str): the uniprot ID Returns: str: amino acid sequence

Adopt cost-sensitive cross-entropy as the default Args: epoch: int. starting from 1.

Adopt the class-balanced loss as default re-weighting method in drw according to LDAM official codes. Args: epoch: int

Args: epoch: int