Diff computes the following derivative: (u - v)' = u' - v'

Diff computes the following derivative: (u + v)' = u' + v'

Diff computes the following derivative: x' = 1. However if i not equal to the Var's Index then a 0 Const is returned.

Diff computes the following derivative: (uv)' = u'v + uv'

Diff compute the following derivative: (1 / u)' = -u' / u².

Diff compute the following derivative: (1 / u)' = -u' / u².

Diff compute the following derivative: |u|' = uu' / |u|.

Diff computes the following derivative: (u / v)' = (u'v - uv') / v²

Eval computes the opposite of each value.

Eval computes the square of each value.

Eval computes the cosine of each value.

Eval computes the sine of each value.

Eval takes the minimum along aligned values.

Eval returns the same value.

Eval takes the maximum along aligned values.

Eval computes the difference between Left and Right.

Eval sums aligned values.

Eval returns the ith column of a matrix.

Eval multiplies aligned values.

Eval computes the inverse of each value.

Eval computes

Eval computes the absolute value of each value.

Eval divides aligned values.

From https://stats.stackexchange.com/questions/224140/step-by-step-example-of-reverse-mode-automatic-differentiation.

Execute RootCmd and catch error.

GetVars returns the Vars contained in an Operator. Pre-order traversal is used.

Gradients computes sign(yPred[i] - yTrue[i]).

Gradients computes 2 * (yPred[i] - yTrue[i]).

Gradients computes yPred[i] - yTrue[i].

MarshalJSON serializes a GradientBoosting.

Name of Neg is "neg".

Name of Square is "square".

Name of Cos is "cos".

Name of Sin is "sin".

Name of Min is "min".

Name of Max is "max".

Name of Sub is "sub".

Name of Add is "add".

Name of Mul is "mul".

Name of Inv is "inv".

Name of If is "if".

Name of Abs is "abs".

Name of Div is "div".

NeedsProbabilities method of MAE.

NeedsProbabilities method of Accuracy.

NeedsProbabilities method of R2.

NeedsProbabilities method of Precision.

NeedsProbabilities method of MicroPrecision.

NeedsProbabilities method of MacroPrecision.

NeedsProbabilities method of WeightedPrecision.

NeedsProbabilities method of RMSE.

NeedsProbabilities method of MSE.

NeedsProbabilities method of Negative.

NeedsProbabilities method of ROCAUC.

NeedsProbabilities method of Recall.

NeedsProbabilities method of MicroRecall.

NeedsProbabilities method of MacroRecall.

NeedsProbabilities method of WeightedRecall.

NeedsProbabilities method of AbsolutePearson.

NeedsProbabilities method of F1.

NeedsProbabilities method of MicroF1.

NeedsProbabilities method of MacroF1.

NeedsProbabilities method of WeightedF1.

NeedsProbabilities method of LogLoss.

NewGradientBoosting returns a GradientBoosting.

Operand returns Neg's operand or nil.

Operand returns Square's operand or nil.

Operand returns Cos's operand or nil.

Operand returns Sin's operand or nil.

Operand returns one of Min's operands, or nil.

Operand returns nil.

Operand returns one of Max's operands, or nil.

Operand returns one of Sub's operands, or nil.

Operand returns one of Add's operands, or nil.

Operand returns nil.

Operand returns one of Mul's operands, or nil.

Operand returns Inv's operand or nil.

Operand returns If's operand or nil.

Operand returns Abs's operand or nil.

Operand returns one of Div's operands, or nil.

ParseFuncs parses a string into a slice of Operators.

ParseMetric returns a Metric from it's String representation.

Predict makes predictions with the best obtained Program as so far.

PredictPartial is a convenience function on top of Predict to make predictions on a single instance.

SetOperand replaces Neg's operand if i is equal to 0.

SetOperand replaces Square's operand if i is equal to 0.

SetOperand replaces Cos's operand if i is equal to 0.

SetOperand replaces Sin's operand if i is equal to 0.

SetOperand replaces one of Min's operands if i is equal to 0 or 1.

SetOperand returns the Const without any modifications.

SetOperand replaces one of Max's operands if i is equal to 0 or 1.