Correctness, which has to be considered at various levels:

• Routine manipulations
• Derivation from fundamentals
• Errors are catastrophic

Efficiency:

• Mathematical transformations to optimize
• Code optimization
• Code integration
• Programming time saved

Numerical stability, detection and reporting of numerical problems, singularities, discontinuities, etc.

Increased reliability by being able to solve the same problem in two completely different environments. (Generation loop).

EXAMPLES

• Automatic code generation (van der Waals)
• Visual verification and Potential computation
• Potential gradient and hessian
• Code optimization, size comparison
• Sample of optimized code
• Relabelling and reusing temporary variables
• Symbolic execution and equivalence testing
• Portion of the real program to generate potential and derivatives
• Real potential and gradient code
• Fitting the internal lengths by least squares
• ditto but for 4 subtrees
• Computing and selecting the best permutation
• Sequence alignment with affine deletion cost