NDS_counter function

def NDS_counter(N,D,dist):
    import numpy as np
    #N = number of observations
    #D = number of dimensions
    #dist = 0 for uniform distribution
    #dist = 1 for normal distribution
    
    
    #FOR UNIFORM DISTRIBUTION
    if dist == 0:    
        solutions = np.zeros((N,D))
        for d in range(0,D):
            solutions[:,d] = np.random.random(N)
        
    #FOR NORMAL DRIBUTION
    if dist == 1:    
        solutions = np.zeros((N,D))
        for d in range(0,D):
            solutions[:,d] = np.random.normal(loc=0.0, scale=1.0, size=N)
        
    
    #count non-dominated solutions 
    #initial = np.zeros((1,D))
    non_dominated_set = np.reshape(solutions[0,:], ((1,D)))
    for i in range(1,N):
        comparison = solutions[i,:] > non_dominated_set
        if all(np.sum(comparison,axis=1)>=1):
            #improves on at least one objective for all solutions  
            #completely better
            inf_loc = [np.squeeze(np.where(np.sum(comparison,axis=1)==D))]
            if len(inf_loc):
                non_dominated_set = np.delete(non_dominated_set, inf_loc,0)
            #add to non dom set
            non_dominated_set = np.concatenate((non_dominated_set, [solutions[i,:]]), axis=0)
    
    return(non_dominated_set, len(non_dominated_set))