numpy交叉验证集的简单构建(Simple construction of numpy cross validation set)

num_folds = 5
k_choices = [1, 3, 5, 8, 10, 12, 15, 20, 50, 100]
# k_choices = [1, 5]

x_train_folds = []
y_train_folds = []

# x_train_folds = np.array(np.split(x_train, num_folds))
# y_train_folds = np.array(np.split(y_train, num_folds))
x_train_folds = np.array(np.split(x_test, num_folds))
y_train_folds = np.array(np.split(y_test, num_folds))

k_to_accuracies = {}
for k in k_choices:
    curr_acc = []
    for i in np.arange(num_folds):
        indx = np.array([j for j in range(num_folds) if j != i])
        X_test_n = x_train_folds[i]
        y_test_n = y_train_folds[i].reshape(-1,)
        x_train_folds[indx]
        x_train_n = np.concatenate(x_train_folds[indx], axis=0)
        y_train_n = np.concatenate(y_train_folds[indx], axis=None)

        classifier = KNearestNeighbor()
        classifier.train(x_train_n, y_train_n)
        dists = classifier.compute_distances_two_loops(X_test_n)
        y_test_n_pred = classifier.predict_labels(dists, k)
        num_correct = np.sum(y_test_n_pred == y_test_n)
        accuracy = float(num_correct) / len(y_test_n)
        curr_acc.append(accuracy)
    k_to_accuracies[k] = curr_acc
————————
num_folds = 5
k_choices = [1, 3, 5, 8, 10, 12, 15, 20, 50, 100]
# k_choices = [1, 5]

x_train_folds = []
y_train_folds = []

# x_train_folds = np.array(np.split(x_train, num_folds))
# y_train_folds = np.array(np.split(y_train, num_folds))
x_train_folds = np.array(np.split(x_test, num_folds))
y_train_folds = np.array(np.split(y_test, num_folds))

k_to_accuracies = {}
for k in k_choices:
    curr_acc = []
    for i in np.arange(num_folds):
        indx = np.array([j for j in range(num_folds) if j != i])
        X_test_n = x_train_folds[i]
        y_test_n = y_train_folds[i].reshape(-1,)
        x_train_folds[indx]
        x_train_n = np.concatenate(x_train_folds[indx], axis=0)
        y_train_n = np.concatenate(y_train_folds[indx], axis=None)

        classifier = KNearestNeighbor()
        classifier.train(x_train_n, y_train_n)
        dists = classifier.compute_distances_two_loops(X_test_n)
        y_test_n_pred = classifier.predict_labels(dists, k)
        num_correct = np.sum(y_test_n_pred == y_test_n)
        accuracy = float(num_correct) / len(y_test_n)
        curr_acc.append(accuracy)
    k_to_accuracies[k] = curr_acc