nmmso_benchmark: CEC Benchmarking Tool

# source("./R/niching_funcs.R")
# source("./R/feval.R")
# source("./R/count_goptima.R")

# # Demonstration file on how to use the benchmark suite of the Competition

# dims = c(1, 1, 1, 2, 2, 2, 2, 3, 3, 2, 2, 2, 2, 3, 3, 5, 5, 10, 10, 20) # dimensionality of benchmark functions
# max_fes = c(50000 * matrix(1, 1, 5), 200000, 200000, 400000, 400000, 200000 * matrix(1, 1, 4), 400000 * matrix(1, 1, 7))

# # do not forget
# initial_flag = 0 # the global flag used in test suite
# for(func_num in 1:20) {
# 	# set the lower and upper bound for each function
# 	# do not forget
# 	initial_flag = 0 # should set the flag to 0 for each run, each function

# 	# dimension of the problem
# 	d = dims[func_num]

# 	# potential solution
# 	x = matrix(1, 1, d)
# 	# evaluate the solution
# 	val = niching_func(x, func_num) # fitness evaluation
# 	#vals = c(vals, val)
# 	cat("f_", func_num, " : f(1...1) = ", val, "\n")
# }

# fgoptima = c(200.0, 1.0, 1.0, 200.0, 1.03163, 186.731, 1.0, 2709.0935, 1.0, -2.0, matrix(0, 1, 10))
# np = 100
# for(func_num in 1:20) {
# 	# do not forget
# 	initial_flag = 0 # should set the flag to 0 for each run, each function
# 	d = dims[func_num]

# 	# randomize population within optimization bounds
# 	# (here dummy initialization within [0, 1] only for demo)
# 	pop = matrix(runif(np * d), np, d)

# 	# how many global optima have been found?
# 	accuracy = 0.001
# 	result = count_goptima(pop, func_num, accuracy)
# 	count = result$count
# 	goptima_found = result$finalseeds
# 	cat("f_", func_num, ", In the current population there are ", count, " global optima.\n")

# 	# print some stuff :-)
# 	if(count != 0) {
# 		for(i in 1:size(goptima_found)[1]) {
# 			val = niching_func(goptima_found[i], func_num)
# 			cat("F_p: ", val, ", F_g: ", fgoptima[func_num], ", diff: ", abs(val - fgoptima[func_num]), ".\n")
# 			cat("F_p - F_g <= ", accuracy, " : ", abs(val - fgoptima[func_num]) < accuracy, ".\n")
# 		}
# 	}
# }