Load cached dataset. :param training: Load training part if training is True, test part otherwise.

Load cached dataset. :param training: Load training part if training is True, test part otherwise.

Load cached dataset. :param training: Load training part if training is True, test part otherwise.

Re-group scores respecting M4 rules. :param scores: Scores per group. :return: Grouped scores.

Re-group scores respecting M4 rules. :param scores: Scores per group. :return: Grouped scores.

Re-group scores respecting M4 rules. :param scores: Scores per group. :return: Grouped scores.

Build mini-batch tensors from a list of (X, mask) tuples. Mask input. Create Args: data: len(batch_size) list of tuples (X, y).

a/b where the resulted NaN or Inf are replaced by 0.

a/b where the resulted NaN or Inf are replaced by 0.

a/b where the resulted NaN or Inf are replaced by 0.

get modes on frequency domain: 'random' means sampling randomly; 'else' means sampling the lowest modes;

Args: df: input dataframe Returns: df: normalized dataframe

Args: df: input dataframe Returns: df: normalized dataframe

Adds a list of values as a bottom row of a table in a given excel file

Randomly create a boolean mask of length `L`, consisting of subsequences of average length lm, masking with 0s a `masking_ratio` proportion o

Returns the squared L2 norm of output layer of given model

The last window of insample size of all timeseries. This function does not support batching and does not reshuffle timeseries.

Used to mask padded positions: creates a (batch_size, max_len) boolean mask from a tensor of sequence lengths, where 1 means keep element at

For the task specified in the configuration returns the corresponding combination of Dataset class, collate function and Runner class.

Plot confusion matrix in a separate window

Print confusion matrix as text to terminal

Reads a single .csv, which typically contains a day of datasets of various weld sessions.