niteshade.utils

General utility and helper functions.

Functions

check_batch_size(y)	Check the batch size of input label data.
check_num_of_classes(y)	Check the number of classes in one hot encoded data.
copy(array_like)	Used to copy a torch.Tensor or np.ndarray object.
decode_one_hot(y)	Decode one hot encoded data.
get_cmap(nlabels)
get_time_stamp_as_string()	Get the current time stamp as a string.
load_model(filename)	Load a binary file containing a neural network.
one_hot_encoding(y, num_classes)	Perform one hot encoding of previiously decoded data.
rand_cmap(nlabels)	Creates a random colormap to be used together with matplotlib.
save_model(model, filename)	Save an object as a binary .pickle file.
save_pickle(results[, dirname, filename])	Save results as a .pickle file.
save_plot(plt[, dirname, plotname])	Save a matplotlib.pyplot.plot() as a .png file.
train_test_MNIST([dir, transform, val_size])	Function to load torchivisions' MNIST dataset, splitted into train, test, and validation sets (the latter only if val_size != None).
train_test_cifar([dir, transform, val_size])	Function to load torchvisions' CIFAR10 dataset, splitted into train, test, and validation sets (the latter only if val_size != None).
train_test_iris([test_size, val_size, ...])	Loads the Iris dataset using sklearn.datasets.load_iris() and returns the inputs and labels in splitted train and test sets (and validation too if val_size != None).