Add numeric diff support for Python CostFunction bindings + example updates

[mrawding]

This PR adds numeric-diff support for Python-defined cost functions so users can skip manual Jacobians, aligning behavior more closely with Ceres’ autodiff-focused docs. It also refreshes the Python examples
(hello world, Powell’s, and curve fitting) and adds an autodiff-style Powell’s + Curve Fitting example that mirrors the Ceres documentation flow.

Core NumericDiff Update

pyrceres.CostFunction now has member variable set_use_numerif_diff that allows users to set to True for auto differentiation and skip deriving jacobian movement in cost functions

Example from examples/curve_fitting_autodiff.py

lass ExpResidual(pyceres.CostFunction):
    def __init__(self, x, y_obs):
        super().__init__()
        self.x = x
        self.y_obs = y_obs
        self.set_num_residuals(1)
        self.set_parameter_block_sizes([1, 1])
        self.set_use_numeric_diff(True)



    def Evaluate(self, parameters, residuals, jacobians):
        m = parameters[0][0]
        c = parameters[1][0]

        # Model prediction
        y_pred = math.exp(m * self.x + c)

        # Residual: r = y_pred - y_obs
        residuals[0] = y_pred - self.y_obs
# Comments Show Math Needed To Complete Problem Without NumericDiff
        # if jacobians is not None:
        #     # ∂r/∂m = x * exp(m x + c)
        #     if jacobians[0] is not None:
        #         jacobians[0][0] = self.x * y_pred

        #     # ∂r/∂c = exp(m x + c)
        #     if jacobians[1] is not None:
        #         jacobians[1][0] = y_pred

        return True

examples updates

• hello.py (solves 1/2(10-x)^2
• curve_fitting.py (solves e^mx+c)
• robust_curve_fitting_loss.py (solves curve_fitting equation with random outliers)
• powells.py
• powells_autodiff.py
• curve_fitting_autodiff.py