def resolve_all_option_collections(self, info, **kwargs):
field_map = {
"option": ("option", "select"),
}
return helpers.optimize(OptionCollection.objects.all(),
ast_to_dict(info.field_asts),
field_map)
def resolve_jobs(self, args, context, info):
field_map = {
"buildPlatform": ("build_platform", "select"),
"jobLog": ("job_log", "prefetch"),
"jobType": ("job_type", "select"),
"jobGroup": ("job_group", "select"),
"failureClassification": ("failure_classification", "prefetch"),
"failureLine": ("job_log__failure_line", "prefetch"),
"group": ("job_log__failure_line__group", "prefetch"),
"textLogStep": ("text_log_step", "prefetch"),
"errors": ("text_log_step__errors", "prefetch"),
}
return helpers.optimize(Job.objects.filter(push=self, **args),
ast_to_dict(info.field_asts), field_map)
def resolve_jobs(self, info, **kwargs):
field_map = {
"buildPlatform": ("build_platform", "select"),
"jobLog": ("job_log", "prefetch"),
"jobType": ("job_type", "select"),
"jobGroup": ("job_group", "select"),
"failureClassification": ("failure_classification", "prefetch"),
"failureLine": ("job_log__failure_line", "prefetch"),
"group": ("job_log__failure_line__group", "prefetch"),
"textLogStep": ("text_log_step", "prefetch"),
"errors": ("text_log_step__errors", "prefetch"),
}
return helpers.optimize(Job.objects.filter(push=self, **kwargs),
ast_to_dict(info.field_asts),
field_map)
def model(X_train,
Y_train,
X_test,
Y_test,
num_iter=2000,
learn_rate=0.5,
print_cost=False):
## PARAMETERS INIT ##
w, b = init_with_zeros(X_train.shape[0])
## GRADIENT DESCENT ##
params, grads, costs = optimize(w, b, X_train, Y_train, num_iter,
learn_rate, print_cost)
## RETRIEVE PARAMS ##
w = params["w"]
b = params["b"]
## PREDICTION ##
Y_predict_test = predict(w, b, X_test)
Y_predict_train = predict(w, b, X_train)
print("train accuracy: {} %".format(
100 - np.mean(np.abs(Y_predict_train - Y_train)) * 100))
print(
"test accuracy: {} %".format(100 -
np.mean(np.abs(Y_predict_test - Y_test)) *
100))
d = {
"costs": costs,
"Y_predict_test": Y_predict_test,
"Y_predict_train": Y_predict_train,
"w": w,
"b": b,
"learn_rate": learn_rate,
"num_iter": num_iter
}
return d
best_space = solution[1]
best_items = solution[2]
print('Best solution exercise A:')
print('Wasted weight: ' + str(best_weight) + ', Wasted space: ' + str(best_space) + ', Unloaded cargo: ' + str(best_items) + '.')
print('')
# # opdracht B (Cargolist 1)
# # loading cargo sorted by density
sortedlist = sorted(cargolist, key = lambda x: [cargolist[x][2]])
helpers.ratio(cargolist, sortedlist, spacecrafts)
solution = helpers.solution(spacecrafts, items)
print('The solution found by sorting the cargo by density for exercise B is as follows:')
print('Wasted weight: ' + str(solution[0]) + ', Wasted space: ' + str(solution[1]) + ', Unloaded cargo: ' + str(solution[2]) + '.')
# # partly brute force on shortened cargolist
optimized = helpers.optimize(cargolist, spacecrafts)
bestscore = 10000; bestsolution = []
for bruteforce in range(0, 100000):
randomlist = optimized.keys()
random.shuffle(randomlist)
helpers.ratio(optimized, randomlist, spacecrafts)
solution = helpers.solution(spacecrafts, items)
# score function
score = solution[0] + 0.01 * solution[2]
# remember best solution
if score < bestscore:
bestscore = score
bestsolution = solution
print("Best solution exercise B:")
print('Score: ' + str(bestscore))