This is a small lib that tries to emulate the famous spreadsheet function that keeps trying a bunch of values until one satisfies your conditions. Very handy to brute force some math problems.
GoalSeeker.seek(
start, # First value to be tried
goal, # Value that your stop function should reach to stop
max_cycles, # Maximum number of iterations that you are willing to spend looking for the answer
function, # fit function, it receives a number and produces another one that will be matched against the goal value
seeker_type # [Optional] Defines which algorithm will be used to search for the goal value. Can be:
# :binary (default): Will use a binary search to find the best fit
# :brute_force: Will look sequentially for the best fit. This algorithm also receives a `step` value that defines the increment of each step, Default is set to 1.
# :genetic: Will use a very simple genetic algorithm to solve find the best fit. There's an optional `epsilon` argument that can be used to specify how accurate you expect the result to be. Default is set to 0. [WARNING] this is a probabilistic strategy, so the timing to solve things can vary. So, use with care.
)Examples:
GoalSeeker.seek(
start: 100,
goal: 0,
max_cycles:100,
function: lambda { |x| x*x -5*x + 6 }
) # Will return 3 (the other solution to the equation is 2 =] )Running tests
Just go and rake test