CODEWITHSUNDEEP
pythonkeras
vincent
vincent

Reputation: 1678

Get the accuracy of model on prediciton

l want to get the accuracy of my model predicting the labels of x_test

   from __future__ import print_function
    from keras.models import Sequential
    from keras.layers import Dense
    import keras
    import numpy as np
    model = Sequential()
    model.add(Dense(2000, input_dim=3072, activation='relu'))
    model.add(Dense(500, activation='relu'))
    model.add(Dense(66, activation='softmax'))
    model.fit(x_train,y_train, epochs=100, batch_size=128)
    scores = model.evaluate(x_train, y_train)
    print("\n%s: %.2f%%" % (model.metrics_names[1], scores[1]*100))

Now l want to get accuracy on prediction 

predictions = model.predict(x_test)

l tried :

  print("\n%s: %.2f%%" % (model.metrics_names[1], predictions*100))

l got the following error:

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-262-edbcf292f31c> in <module>()
----> 1 print("\n%s: %.2f%%" % (model.metrics_names[1], predictions*100))

TypeError: float argument required, not numpy.ndarray

Upvotes: 1

Views: 3113

Answers (1)

Marcin Możejko
Marcin Możejko

Reputation: 40506

model.predict produces a numpy.array which is something completely different from float. You might try to print that using print(predictions) but using formatted string with float absolutely won't work in this case. Try:

print("\n%s:" % (model.metrics_names[1]))
print(100 * predictions)

or 

print("\n%s: %s" % (model.metrics_names[1], np.array_str(predictions*100)))

or if you have only one case in x_test:

print("\n%s: %.2f%%" % (model.metrics_names[1], predictions[0]*100))

Upvotes: 1

Related Questions