RSVP für Ihr lokales TensorFlow Everywhere-Event noch heute!

tfl.visualization.plot_outputs

Visualises several outputs for same set of inputs.

Used in the notebooks

Used in the tutorials

This is generic plotting helper not tied to any layer. Can visualize either:

  • 2-d graphs: 1-d input, 1-d output.
  • 3-d surfaces: 2-d input, 1-d output.

inputs one of:

  • ordered list of 1-d points
  • tuple of exactly 2 elements which represent X and Y coordinates of 2-d mesh grid for pyplot 3-d surface visualization. See test_utils.two_dim_mesh_grid for more details.
outputs_map dictionary {name: outputs} where "outputs" is a list of 1-d points which correspond to "inputs". "name" is an arbitrary string used as legend.
file_path if set - visualisation will be saved as png at specified location.
figsize The figsize parameter passed to pyplot.figure().

ValueError if configured to visualise more than 4 3-d plots.

Pyplot object containing visualisation.