Live.log_plot()
Generates a
scikit learn plot and
saves the data in {path}/plots/{name}.json.
def log_plot(self, name: str, labels, predictions, **kwargs):Usage:
from dvclive import Live
live = Live()
y_true = [0, 0, 1, 1]
y_pred = [1, 0, 1, 0]
y_score = [0.1, 0.4, 0.35, 0.8]]
live.log_plot("roc", y_true, y_score)
live.log_plot("confusion_matrix", y_true, y_pred)Description
Uses name to determine which plot should be generated. See
supported plots.
The generated {path}/plots/{name}.json can be visualized with dvc plots.
Step updates
Live.log_plot() can be currently only used when step is None.
If you perform step updates in your code, you can later use
Live.set_step(None) in order to be able to use Live.log_plot().
for epoch in range(NUM_EPOCHS):
live.log(metric_name, value)
live.next_step()
live.set_step(None)
live.log_plot("roc", y_true, y_score)Supported plots
name must be one of the supported plots:
Generates a calibration curve plot.
Calls
sklearn.calibration.calibration_curve
and stores the data at {path}/plots/calibratrion.json in a format compatible
with dvc plots.
y_true = [0, 0, 1, 1]
y_score = [0.1, 0.4, 0.35, 0.8]
live.log_plot("calibration", y_true, y_score)Example usage with dvc plots:
$ dvc plots show 'dvclive/plots/calibration.json' \
-x prob_pred -y prob_true \
--x-label 'Mean Predicted Probability' \
--y-label 'Fraction of Positives' \
--title 'Calibration Curve'
Generates a confusion matrix plot.
Stores the abels and predictions in {path}/plots/confusion_matrix.json, with
the format expected by the confusion matrix
template of
dvc plots.
y_true = [1, 1, 2, 2]
y_pred = [2, 1, 1, 2]
live.log_plot("confusion_matrix", y_true, y_pred)Example usage with dvc plots:
$ dvc plots show 'dvclive/plots/confusion_matrix.json' \
-x actual -y predicted \
--template confusion
Generates a detection error tradeoff (DET) plot.
Calls
sklearn.metrics.det_curve
and stores the data at {path}/plots/det.json in a format compatible with
dvc plots.
y_true = [1, 1, 2, 2]
y_score = [0.1, 0.4, 0.35, 0.8]
live.log_plot("det", y_true, y_score)Example usage with dvc plots:
$ dvc plots show 'dvclive/plots/det.json' \
-x fpr -y fnr \
--title 'DET Curve'
Generates a precision-recall curve plot.
Calls
sklearn.metrics.precision_recall_curve
and stores the data at {path}/plots/precision_recall.json in a format
compatible with dvc plots.
y_true = [1, 1, 2, 2]
y_score = [0.1, 0.4, 0.35, 0.8]
live.log_plot("precision_recall", y_true, y_score)Example usage with dvc plots:
$ dvc plots show 'dvclive/plots/precision_recall.json' \
-x recall -y precision \
--title 'Precision Recall Curve'
Generates a receiver operating characteristic (ROC) curve plot.
Calls
sklearn.metrics.roc_curve
and stores the data at {path}/plots/roc.json in a format compatible with
dvc plots.
y_true = [1, 1, 2, 2]
y_score = [0.1, 0.4, 0.35, 0.8]
live.log_plot("roc", y_true, y_score)Example usage with dvc plots:
$ dvc plots show 'dvclive/plots/roc.json' \
-x fpr -y tpr \
--title 'ROC Curve'
Parameters
-
name- a supported plot type -
labels- array of ground truth labels -
predictions- array of predicted labels (forconfusion_matrix) or predicted probabilities (for other plots) -
**kwargs- additional arguments to be passed to the internal scikit-learn function being called
Exceptions
-
dvclive.error.InvalidPlotTypeError- thrown if the providednamedoes not correspond to any of the supported plots. -
RuntimeError- thrown ifLive.log_plot()is used andstepis notNone.