Package index • vistool

All functions

hypothesis(): Hypothesis wrapper for functional models

LossFunction: Loss Function

Objective: Objective function

Optimizer: Optimizer class

OptimizerGD: Gradient descent optimizer

OptimizerMomentum: Momentum optimizer

OptimizerNAG: Nesterovs momentum optimizer

Visualizer: Base Visualizer Class

VisualizerLossFuns: Visualizer for Loss Functions

VisualizerModel: Visualize Model (Unified 1D/2D)

VisualizerObj: Visualize Objective (Unified 1D/2D)

VisualizerSurface: Visualize 2D Functions as Interactive Surfaces

VisualizerSurfaceModel: Visualize Model as Interactive Surface

VisualizerSurfaceObj: Visualize Objective as Interactive Surface

as.data.table(<DictionaryLoss>): Convert Dictionary to Data Table

as_visualizer(): Convert to visualizer

assertStepSizeControl(): Assertion for the main signature of a `stepSizeControlXX` function.

dict_loss: Dictionary of Loss Functions

dict_objective: Dictionary for test functions

get_continuous_colorscale(): Get continuous colorscale for surface plots

get_vistool_color(): Get consistent color from discrete palette

lss(): Retrieve Loss Function

mergeOptimArchives(): Merge optimization archives

mlr_learners_regr.lm_formula LearnerRegrLMFormula: Linear Model Regression Learner with Formula

obj(): Retrieve Objective Functions

stepSizeControlDecayExp(): Conduct exponential decay to adjust the update. See https://neptune.ai/blog/how-to-choose-a-learning-rate-scheduler

stepSizeControlDecayLinear(): Conduct linear decay to adjust the update. See https://neptune.ai/blog/how-to-choose-a-learning-rate-scheduler

stepSizeControlDecaySteps(): Conduct a step-wise decay to adjust the update. See https://neptune.ai/blog/how-to-choose-a-learning-rate-scheduler

stepSizeControlDecayTime(): Conduct time decay to adjust the update. See https://neptune.ai/blog/how-to-choose-a-learning-rate-scheduler

stepSizeControlLineSearch(): Conduct line search in each iteration to adjust the update.

vistool_theme(): vistool theming utilities