I realize that you may not be maintaining this algorithm anymore, but as it is a (if not, the) main alternative to ABAGAIL for ML students, it would be beneficial to take on this capability. Thanks in advance! :)
Changes needed:

1. Add interations to the init function:
   def init(self, domain, fitness_function, samples=1000, iterations=100,
   percentile=0.90, typeDomain=None):

2. Save this and make it available within the function:
   self.m = iterations

3. Adjust the fit() function to iterate:
   def fit(self):
   """
   Run this to perform m iterations of the Mimic algorithm
   :return: A list containing the top percentile of data points
   """
   for i in range(self.m):
   samples = self.sample_set.get_percentile(self.percentile)
   self.distribution = Distribution(samples)
   self.sample_set = SampleSet(
   self.distribution.generate_samples(self.samples),
   self.fitness_function)
   return self.sample_set.get_percentile(self.percentile)

mimcry.mimic.Mimic.fit() returning TypeError: fit() takes exactly 1 arguments (2 given) even though only one is given. Please advise.

Occasionally this happens: https://travis-ci.org/mjs2600/mimicry/builds/52506167 We should probably figure out why.