The test suite includes a simple recursive Fibonacci sequence generator, but it's terribly inefficient. One simple method for improving its performance is using a technique called memoization. Write a function that takes a function_handle that we wish to memoize and returns a function_handle to a memoized version of the initial handle.
Whoops, accidentally had the answer in the test suite. Fixed now.
Bryant,
Any chance you can elaborate a bit? I can make a memoized fibonacci code, but that's not really what you're asking. I don't get what the function call with 2 arguments (and 2nd argument the function itself) is in the test set: seq = memfib(n-1, memfib);
Sorry for the confusion.
Paul, it uses a feature that was just introduced in R2017a. If the answer hadn't accidentally been in the test suite, I never would have figured this one out either!
I tried comparing this code with a non-memoized version, and it was 6 times slower!
Love the name of the problem.^^
Back to basics 8 - Matrix Diagonals
680 Solvers
628 Solvers
307 Solvers
Permute diagonal and antidiagonal
166 Solvers
179 Solvers