|
@@ 123-145 (lines=23) @@
|
| 120 |
|
|
| 121 |
|
return best_chromosome, best_ff |
| 122 |
|
|
| 123 |
|
|
| 124 |
|
def get_clusters(self): |
| 125 |
|
"""! |
| 126 |
|
@brief Returns list of allocated clusters, each cluster contains indexes of objects from the data. |
| 127 |
|
|
| 128 |
|
@return (list) List of allocated clusters. |
| 129 |
|
|
| 130 |
|
@see process() |
| 131 |
|
|
| 132 |
|
""" |
| 133 |
|
|
| 134 |
|
raise NameError("Implementation is require."); |
| 135 |
|
|
| 136 |
|
|
| 137 |
|
@staticmethod |
| 138 |
|
def _select(chromosomes, data, count_clusters, select_coeff): |
| 139 |
|
"""! |
| 140 |
|
@brief Performs selection procedure where new chromosomes are calculated. |
| 141 |
|
|
| 142 |
|
@param[in] chromosomes (numpy.array): Chromosomes |
| 143 |
|
|
| 144 |
|
""" |
| 145 |
|
|
| 146 |
|
# Calc centers |
| 147 |
|
centres = ga_math.get_centres(chromosomes, data, count_clusters) |
| 148 |
|
|
|
@@ 147-168 (lines=22) @@
|
| 144 |
|
""" |
| 145 |
|
|
| 146 |
|
# Calc centers |
| 147 |
|
centres = ga_math.get_centres(chromosomes, data, count_clusters) |
| 148 |
|
|
| 149 |
|
# Calc fitness functions |
| 150 |
|
fitness = genetic_algorithm._calc_fitness_function(centres, data, chromosomes) |
| 151 |
|
|
| 152 |
|
for _idx in range(len(fitness)): |
| 153 |
|
fitness[_idx] = math.exp(1 + fitness[_idx] * select_coeff) |
| 154 |
|
|
| 155 |
|
# Calc probability vector |
| 156 |
|
probabilities = ga_math.calc_probability_vector(fitness) |
| 157 |
|
|
| 158 |
|
# Select P chromosomes with probabilities |
| 159 |
|
new_chromosomes = np.zeros(chromosomes.shape, dtype=np.int) |
| 160 |
|
|
| 161 |
|
# Selecting |
| 162 |
|
for _idx in range(len(chromosomes)): |
| 163 |
|
new_chromosomes[_idx] = chromosomes[ga_math.get_uniform(probabilities)] |
| 164 |
|
|
| 165 |
|
return new_chromosomes |
| 166 |
|
|
| 167 |
|
|
| 168 |
|
@staticmethod |
| 169 |
|
def _crossover(chromosomes): |
| 170 |
|
"""! |
| 171 |
|
@brief Crossover procedure. |
annoviko /
pyclustering
