如何重塑我的 nupy 数组，以在 Keras 中进行有效预测

Question

i trained my n-network, and everything work fine, except that i don't know how to format my data to make a prediction on data that are not in training and testing set.

• I loaded csv data.

• I split it into training and testing set, and everything fork fine for

  x_train, x_test, y_train, y_test = model_selection.train_test_split(X, y, test_size=0.1, random_state=0) 
  \n\n
  bestmodel.fit(x_train, y_train, epochs=1, batch_size=5) 

  i got like 97% acc. For

  print(type(x_test)) print(x_test.dtype) print(x_test.shape)

i have output like

class 
'numpy.ndarray'
float64
(905, 14)

i made my own example,

 z = np.array([1, 1, 1, 0, 1, 0, 1, 0, 1, 0, 0, 1, 0, 1]).astype(float)

    np.reshape(z, (14,)) 

but when i try

bestmodel.predict(z)

i got error

raceback (most recent call last): \n  File "/home/administrator/PycharmProjects/BankMarketinData/main.py", line 81, in \n    main()\n  File "/home/administrator/PycharmProjects/BankMarketinData/main.py", line 76, in main\n    score = bestmodel.predict(z)\n  File "/home/administrator/anaconda3/lib/python3.6/site-packages/keras/engine/training.py", line 1149, in predict\n    x, _, _ = self._standardize_user_data(x) \n  File "/home/administrator/anaconda3/lib/python3.6/site-packages/keras/engine/training.py", line 751, in _standardize_user_data\n    exception_prefix='input')\n  File "/home/administrator/anaconda3/lib/python3.6/site-packages/keras/engine/training_utils.py", line 138, in standardize_input_data\n    str(data_shape)) \nValueError: Error when checking input: expected dense_1_input to have shape (14,) but got array with shape (1,) 

Can you help me reshape and format this z table, that i can use it for prediciton ?

Answer 1

You need to add the batch dimension with a value of 1:

z = np.array([1, 1, 1, 0, 1, 0, 1, 0, 1, 0, 0, 1, 0, 1]).astype(float)
# z.shape is (14,)
z = np.expand_dims(z, axis=0)
# z.shape is now (1, 14)

bestmodel.predict(z)