Originally reported byLuca Geretti (Bitbucket: lgeretti, GitHub: lgeretti)
Given a step h and a grid length k, we should have that the limit of k/h goes to zero for infinite accuracy. This means that the grid length must reduce quicker than the step, for example with k ~ h^2.
Originally reported by Luca Geretti (Bitbucket: lgeretti, GitHub: lgeretti)
Given a step h and a grid length k, we should have that the limit of k/h goes to zero for infinite accuracy. This means that the grid length must reduce quicker than the step, for example with k ~ h^2.