Method: wandb.plot.roc()

Log an ROC curve in one line:

wandb.log({"roc" : wandb.plots.roc(ground_truth, predictions, class_names=None)})

You can log this whenever your code has access to:

• a model's predicted scores (predictions) on a set of examples
• the corresponding ground truth labels (ground_truth) for those examples
• (optionally) a list of the class label names (class_names=["cat", "dog", "bird"...] if label index 0 means cat, 1 = dog, 2 = bird, etc.)

Basic usage

I finetune a CNN to predict 10 classes of living things: plants, birds, insects, etc. In my validation step, I call

wandb.log({"my_custom_plot_id" : wandb.plots.roc(ground_truth,
                        predictions, class_names=["Amphibia", "Animalia", ... "Reptilia"])

to produce the following curve for each run of my model (where each run logs to the same plot key, my_custom_plot_id):

Customized usage

To make this chart more legible, I can simply edit the built in wandb chart definition (or Vega spec), following the Vega visualization grammar and produce this new chart, where the differences between a single class are much easier to spot across runs: