Overview
Machine learning plotting functions provide visualizations for model evaluation and diagnostics.
Classification Visualizations
plotConfusionMatrix
Plot a confusion matrix as a heatmap.
function plotConfusionMatrix(
cm: Tensor,
labels?: readonly string[],
options?: PlotOptions
): void
Confusion matrix tensor (2D)
Class labels for axis annotations
import { plotConfusionMatrix, confusionMatrix, tensor } from 'deepbox';
const yTrue = tensor([0, 1, 1, 0, 1, 2, 2, 0]);
const yPred = tensor([0, 1, 0, 0, 1, 2, 1, 0]);
const cm = confusionMatrix(yTrue, yPred);
plotConfusionMatrix(cm, ['Class 0', 'Class 1', 'Class 2']);
plotRocCurve
Plot a ROC curve with optional AUC annotation.
function plotRocCurve(
fpr: Tensor,
tpr: Tensor,
auc?: number,
options?: PlotOptions
): void
Area under curve (optional, for title annotation)
import { plotRocCurve, rocCurve, rocAucScore, tensor } from 'deepbox';
const yTrue = tensor([0, 0, 1, 1]);
const yScore = tensor([0.1, 0.4, 0.35, 0.8]);
const [fpr, tpr, _] = rocCurve(yTrue, yScore);
const auc = rocAucScore(yTrue, yScore);
plotRocCurve(fpr, tpr, auc);
plotPrecisionRecallCurve
Plot a precision-recall curve with optional AP annotation.
function plotPrecisionRecallCurve(
precision: Tensor,
recall: Tensor,
averagePrecision?: number,
options?: PlotOptions
): void
Average precision score (optional, for title annotation)
import { plotPrecisionRecallCurve, precisionRecallCurve, averagePrecisionScore, tensor } from 'deepbox';
const yTrue = tensor([0, 0, 1, 1]);
const yScore = tensor([0.1, 0.4, 0.35, 0.8]);
const [prec, rec, _] = precisionRecallCurve(yTrue, yScore);
const ap = averagePrecisionScore(yTrue, yScore);
plotPrecisionRecallCurve(prec, rec, ap);
plotDecisionBoundary
Plot a classifier decision boundary on a 2D feature space.
function plotDecisionBoundary(
X: Tensor,
y: Tensor,
model: { predict: (x: Tensor) => Tensor },
options?: PlotOptions
): void
Feature matrix of shape [n_samples, 2]
Class labels of shape [n_samples]
Model with a predict method that accepts a tensor and returns predictions
Plot styling options (colormap, colors, size, etc.)
import { plotDecisionBoundary, tensor } from 'deepbox';
// Assuming you have a trained model
const X = tensor([[1, 2], [2, 3], [3, 1], [5, 6]]);
const y = tensor([0, 0, 1, 1]);
plotDecisionBoundary(X, y, trainedModel);
Training Visualizations
plotLearningCurve
Plot training and validation learning curves.
function plotLearningCurve(
trainSizes: Tensor,
trainScores: Tensor,
valScores: Tensor,
options?: PlotOptions
): void
Training scores at each size
Validation scores at each size
Plot styling options (colors, etc.)
import { plotLearningCurve, tensor } from 'deepbox';
const trainSizes = tensor([100, 200, 300, 400, 500]);
const trainScores = tensor([0.85, 0.87, 0.88, 0.89, 0.90]);
const valScores = tensor([0.75, 0.78, 0.80, 0.81, 0.82]);
plotLearningCurve(trainSizes, trainScores, valScores);
plotValidationCurve
Plot training and validation curves across hyperparameter values.
function plotValidationCurve(
paramRange: Tensor,
trainScores: Tensor,
valScores: Tensor,
options?: PlotOptions
): void
Training scores for each parameter value
Validation scores for each parameter value
import { plotValidationCurve, tensor } from 'deepbox';
const alphaValues = tensor([0.001, 0.01, 0.1, 1.0, 10.0]);
const trainScores = tensor([0.92, 0.90, 0.88, 0.85, 0.80]);
const valScores = tensor([0.78, 0.82, 0.85, 0.84, 0.79]);
plotValidationCurve(alphaValues, trainScores, valScores);