plot_similarity

torchhd.utils.plot_similarity(input: Tensor, memory: Tensor, ax=None, **kwargs)[source]

Plots the similarity of an one hypervector with a set of hypervectors.

Parameters:

  • input (torch.Tensor) – Hypervector to compare against the memory.

  • memory (torch.Tensor) – Set of \(n\) hypervectors.

  • ax (matplotlib.axes, optional) – Axes in which to draw the plot.

  • **kwargs – matplotlib.axes.Axes.stem arguments.

Returns:

matplotlib.container.StemContainer.

Return type:

StemContainer

Shapes:

  • Input: \((d)\)

  • Memory: \((n, d)\)

Examples:

>>>  import matplotlib.pyplot as plt
>>>  hv = torchhd.level(10, 10000)
>>>  utils.plot_similarity(hv[4], hv)
>>>  plt.show()