torchsig.datasets.dataset_utilsΒΆ
Dataset Utilities
Functions
Collates a batch by zipping its elements together. |
|
Generates the full path for a dataset based on its type, impairment level, and whether it is for training. |
|
Randomly shifts the frequency of a signal to a new center frequency and applies aliasing filters if necessary. |
|
Determines if the dataset will generate 'raw' IQ data, which means no transform and target transforms have been applied. |
|
Converts the input dataset metadata to an appropriate DatasetMetadata object. |