Computes query probability given the worlds probability P(w). Args: self: instance query: query worlds_pr

Entropy Args: probabilities (ndarray): probability vector n_values (int): n values Returns: entropy_values: entropy

Evaluate f1 and accuracy Args: true: Groundtruth values pred: Predicted values Returns: ac: accuracy f1: f1

Extracts P(q) contained in the dictionary 'res' resulting from ProbLog model evaluation.

embed_dim: output dimension for each position pos: a list of positions to be encoded: size (M,) out: (M, D)

Function which counts the occurrece and accuracy Args: n_concepts (int): number of concepts c_pred (ndarray): predicted concepts

Returns DataLoader instances for an operation on MNIST images. Args: count_train: Number of training samples to use (max 60000).

Return set of symbol pairs in a word. Word is represented as tuple of symbols (symbols being variable-length strings).

Tests whether the given value is the identity element of the multiplication up to a small constant

Tests whether the given value is the identity element of the addition up to a small constant

KL between two normal distributions Args: mu: mean of the Gaussian logsigma: std of the Gaussian reduction: which reductio

Computes the probability for each ProbLog fact given the latent vector z Args: self: instance z (torch.tensor): laten

Performs rejection sampling to get valid samples of w_0 that optimise for satisfying constraints. Args: p_w_SBWD: Probability dis

Sets the seeds at a certain value. Args: param seed: the value to be set Returns: None: This function does not return a valu

Shuffle the indices for each operand set.

TRAINING Args: model (MnistDPL): network dataset (BaseDataset): dataset Kandinksy _loss (ADDMNIST_DPL): loss function

Compute the mode of a tensor across all dimensions. That is, the most frequently occuring D-dimensional vector, sample-wise Has much high

Updates weights of graph semiring with the facts probability

Variance Args: probabilities (ndarray): probability vector n_values (int): n values Returns: variance_values: varian

Addmnist concept match loss Args: out_dict: output dictionary args: command line arguments Returns: loss: loss value

Addmnist entropy loss Args: out_dict: output dictionary args: command line arguments Returns: loss: loss value

ADDMMNIST evaluation Args: out_dict (Dict[str]): dictionary of outputs concepts: concepts Returns: loss: loss

Addmnist sat agg loss Args: p1: probability of the first concept p2: probability of the second concept la

Kandinsky classification loss Args: out_dict: output dictionary args: command line arguments Returns: loss: loss val

Kandinsky concept match loss Args: out_dict: output dictionary args: command line arguments Returns: loss: loss valu

Kandinsky entropy loss Args: out_dict: output dictionary args: command line arguments Returns: loss: loss value

KAND evaluation Args: out_dict (Dict[str]): dictionary of outputs debug: debug mode Returns: loss: loss cacc

XOR classification loss Args: out_dict: output dictionary args: command line arguments Returns: loss: loss value

XOR concept match loss Args: out_dict: output dictionary args: command line arguments Returns: loss: loss value

XOR entropy loss Args: out_dict: output dictionary args: command line arguments Returns: loss: loss value lo

XOR evaluation Args: out_dict (Dict[str]): dictionary of outputs Returns: loss: loss cacc: concept accuracy

Multioperation sat agg loss Args: eltn: eltn p1: probability of the first concept p2: probability of the second concept

Prodmnist sat agg loss Args: eltn: eltn p1: probability of the first concept p2: probability of the second concept

SDDOIA bce Args: out_dict: output dictionary args: command line arguments Returns: loss: loss value losses:

SDDOIA bce Args: out_dict: output dictionary args: command line arguments Returns: loss: loss value losses:

SDDOIA classification loss Args: out_dict: output dictionary args: command line arguments Returns: loss: loss value

SDDOIA concept match loss Args: out_dict: output dictionary args: command line arguments Returns: loss: loss value

SDDOIA entropy loss Args: out_dict: output dictionary args: command line arguments Returns: loss: loss value

SDDOIA evaluation Args: out_dict (Dict[str]): dictionary of outputs Returns: loss: loss cacc: concept accuracy

SDDOIA sat agg loss Args: eltn: eltn pCs: probability of the concepts labels: labels grade: grade Returns:

XOR classification loss Args: out_dict: output dictionary args: command line arguments Returns: loss: loss value

XOR concept match loss Args: out_dict: output dictionary args: command line arguments Returns: loss: loss value

XOR entropy loss Args: out_dict: output dictionary args: command line arguments Returns: loss: loss value lo

XOR evaluation Args: out_dict (Dict[str]): dictionary of outputs Returns: loss: loss cacc: concept accuracy

save_names: {layer_name:save_path} dict Returns True if there is a file corresponding to each one of the values in save_names, else Retur

Initialize bins for ECE computation Args: num_bins (int): number of bins Returns: bins: dictioanry containing confidence, ac

Generates random indices for each operand set.

Populates the bins for ECE computation Args: confs (ndarray): confidence preds (ndarray): predictions labels (ndarray): l