Development of customizable analytical derivation tool for algebraic function

One of the major problems in differential calculus is determining the derivative value of a function. A commonly used method is numerical derivative. The value obtained by this method is just approximation value. More accurate derivative value is obtained through analytic approach, by looking for general forms of the derivative function. Problem solving agent approach can be utilized to develop such tool, where the operators and constants in the function should be easily customized. It requires four steps in deriving function: parsing, deriving, simplification, and output returning step. User enters the function in a string, and it will be converted into an abstract syntax tree. This tree is traced by DFS algorithm to be derived and simplified. Based on the test results, the analytical derivation requires short processing time. Derivative results are correct, but simplifications are less optimal in some test cases. The modification of constants, operators, and the derivative rules can be done easily.