How to cross validate ML hyperparameters with genetic algorithm PyGAD?

Question

I'm trying to use the Genetic Algorithm library, PyGAD to cross validate hyper-parameters in non-DL machine learning applications.

Data generation

## Create synthetic data
f = np.vectorize(lambda x: int(x>0.5))
X = np.random.uniform(low=0, high=1, size=3000)
X = np.apply_along_axis(f, 0, X).reshape([1000,3])

def gen_y(X):
  if np.sum(X) > 2:
    return 1 
  elif np.sum(X) < 2:
    return 0 
  else:
    return int(np.random.uniform(0,1) > 0.5)  

y = np.apply_along_axis(gen_y, 1, X)
import pandas as pd 
data = pd.DataFrame(data=X, columns=['x1','x2','x3'])
data['y'] = y 

Genetic Algorithm to ID optimal hyper-parameters

gene_space = [ 
    # n_estimators
    np.linspace(50,200,25, dtype='int'),
    # min_samples_split, 
    np.linspace(2,10,5, dtype='int'),
    # min_samples_leaf,
    np.linspace(1,10,5, dtype='int'),
    # min_impurity_decrease
    np.linspace(0,1,10, dtype='float')
]

def fitness_function_factory(data=data, y_name='y', sample_size=100):

    def fitness_function(solution, solution_idx):
        model = RandomForestClassifier(
            n_estimators=solution[0],
            min_samples_split=solution[1],
            min_samples_leaf=solution[2],
            min_impurity_decrease=solution[3]
        )
        
        X = data.drop(columns=[y_name])
        y = data[y_name]
        X_train, X_test, y_train, y_test = train_test_split(X, y,
                                                            test_size=0.5)        

        train_idx = sample_without_replacement(n_population=len(X_train), 
                                              n_samples=sample_size)         
        
        test_idx = sample_without_replacement(n_population=len(X_test), 
                                              n_samples=sample_size) 
         
        model.fit(X_train[train_idx], y_train[train_idx])
        fitness = model.score(X_test[test_idx], y_test[test_idx])
        
        return fitness 

    return fitness_function

cross_validate = pygad.GA(gene_space=gene_space,
                      fitness_func=fitness_function_factory(),
                      num_generations=100,
                      num_parents_mating=2,
                      sol_per_pop=8,
                      num_genes=len(gene_space),
                      parent_selection_type='sss',
                      keep_parents=2,
                      crossover_type="single_point",
                      mutation_type="random",
                      mutation_percent_genes=25)

The last step is identifying the optimal combination via best_solution method.

cross_validate.best_solution()
>>>
KeyError: "None of [Int64Index([119, 342,  34,  80,  94, 270, 443, 468, 401, 133, 400, 362, 455,\n            480, 449, 271, 303, 399, 462, 237, 152, 264, 281, 301, 435, 386,\n             92, 453, 378, 290, 235,  64, 394,  70, 174, 215,  22, 244, 155,\n            207,  74, 147, 178, 267, 347,  97, 396, 292, 120, 375, 113, 169,\n            460,  43, 168, 298,  37, 300,  91, 331, 388, 321, 481,  96, 308,\n            211, 478, 464,   8, 170,  73, 175, 172, 487, 263, 213, 146, 479,\n            336, 346,  67, 160, 277, 397,  38,   7, 247, 128,  47, 428, 454,\n            313, 257, 338, 199, 381,  60, 245, 324, 223],\n           dtype='int64')] are in the [columns]"

So I believe that this error means that optimal configuration cannot be found. But how is this possible and how can this be remedied?