torch.tx.grad

function grad(pattern: string, variable: string): string

Compute the symbolic derivative of a pattern with respect to a variable.

Parameters

patternstring: The expression pattern to differentiate
variablestring: The variable name to differentiate with respect to

Returns

string– A pattern string representing the derivative

Examples

grad('x ** 2', 'x')  // Returns: '2 * x'
grad('sin(x)', 'x')  // Returns: 'cos(x)'
grad('x * y', 'x')   // Returns: 'y'
grad('sigmoid(x)', 'x')  // Returns: 'sigmoid(x) * (1 - sigmoid(x))'