úÎ2ý0Ó      Safe-Inferred  !"#$%&'#$%&'  !"#$%&'NoneOptimizer configuration. tgt is the type of the value to be  optimized. The Function to be optimized. (Extract the parameters to be tuned from tgt. Create a value of type tgt from the parameters. $An initial guess of the parameters. The global scaling factor. ,Typical deviation of each input parameters. 6 The length of the list is adjusted to be the same as 8 initXs, e.g. you can lazily use an infinite list here. 'Typical mean of each input parameters. < The length of this list too, is adjusted to be the same as  initXs. $Assume the function to be rugged and/ or noisy 6How many re-evaluation to make to estimate the noise. :Perturb the parameters by this amount (relative to sigma)  to estimate the noise /Terminate when one of the scaling grew too big " (initial scaling was too small.) 0Terminate when the global scaling grew too big. ;Terminate when the function value diversity in the current 5 and last few generations is smaller than this value ;Terminate when the improvement is not seen for this number  of iterations. ;Terminate when the deviations in the solutions are smaller  than this value. &Repeat the CMA-ES output into stderr. AInterfaces for passing other configuration arguments directly to  cma.py  The default Config, values. Also consult the original document   .http://www.lri.fr/~hansen/pythoncma.html#-fmin for default values $ of the parameters not listed here. ICreate a minimizing problem, given a pure function and an initial guess. &Create a minimizing problem, given an IO function and an initial guess. DCreate a minimizing problem for a function on traversable structure t. QCreate a minimizing problem for an effectful function on a traversable structure t. >Create a minimizing problem for a function on almost any type a which contain Doubles. JCreate a minimizing problem for an effectful function of almost any type. (ASilently check for python version and place the correct shebang ,Execute the optimizer and get the solution. CgetDoubles and putDoubles are generic functions used to put [Double] in and out of generic data types. Let' s test them. 'let d3 = (1,2,3) :: (Double,Int,Double) getDoubles d3 [1.0,3.0]putDoubles [4,5] d3 (4.0,2,5.0)Clet complicated = ([0,1],(2,[3,4])) :: ([Double],(Double,[Double]))getDoubles complicated[0.0,1.0,2.0,3.0,4.0]"putDoubles [5,6,7,8,9] complicated([5.0,6.0],(7.0,[8.0,9.0]))=Putting back the obtained values should not change the data. import Test.DocTest.PropAtype Complicated = ([[Double]],(),(([(Double,String)]),[Double]))Fprop ((\x -> putDoubles (getDoubles x) x == x) :: Complicated -> Bool)5You can get the original list back after putting it. #let make3 xs = take 3 $ xs ++ [0..]wprop ((\xs' y -> let xs = make3 xs' in getDoubles (putDoubles xs y)==xs) :: [Double] -> (Double,Double,Double) -> Bool) ()    ()*      !"#$%&'()*+, cmaes-0.2.1.1%Numeric.Optimization.Algorithms.CMAES Paths_cmaesConfigfuncIO projection embeddinginitXssigma0scaling typicalXs noiseHandling noiseReEvalsnoiseEps tolFacUpX tolUpSigmatolFun tolStagnationtolXverbose otherArgs defaultConfigminimize minimizeIO minimizeT minimizeTIO minimizeG minimizeGIOrun getDoubles putDoublescatchIOversionbindirlibdirdatadir libexecdir getBinDir getLibDir getDataDir getLibexecDirgetDataFileNamewrapperFnFullPathzipTWith