形状不符合我的阵列预期

Question

您好，请考虑以下代码：

# load the mnist training data CSV file into a list
training_data_file = open("Training_Set/mnist_train_100.csv", 'r')
training_data_list = training_data_file.readlines()
training_data_file.close()

caches_x = []
caches_y = []
for record in training_data_list:
    all_values = record.split(',')
    x_inputs = (np.asfarray(all_values[1:]) / 255.0 * 0.99) + 0.01
    y_inputs = (np.asfarray(all_values[0]))

    cache_x = (x_inputs)
    caches_x.append(cache_x)

    cache_y = (y_inputs)
    caches_y.append(cache_y)



train_x = np.array(caches_x).T

print("shape of train_x = " + str(train_x.shape))
print("train_x =" + str(train_x))

train_y = np.array(caches_y).T

print("shape of train_y = " + str(train_y.shape))
print("train_y =" + str(train_y))

当我打印train_x的形状时，我的期望值为（784,103）。

当我打印train_y的形状时，我有（103，）

我做错了什么？同样对于train_y，我希望（或想要拥有）形状（1,103） 有人可以帮我吗？ tnx。