chainer.functions.mean_absolute_error

chainer.functions.mean_absolute_error(x0, x1)[source]

Mean absolute error function.

The function computes the mean absolute error between two variables. The mean is taken over the minibatch. Args x0 and x1 must have the same dimensions. This function first calculates the absolute value differences between the corresponding elements in x0 and x1, and then returns the mean of those differences.

Parameters
Returns

A variable holding an array representing the mean absolute error of two inputs.

Return type

Variable

Example

1D array examples:

>>> x = np.array([1, 2, 3]).astype(np.float32)
>>> y = np.array([0, 0, 0]).astype(np.float32)
>>> F.mean_absolute_error(x, y)
variable(2.)
>>> x = np.array([1, 2, 3, 4, 5, 6]).astype(np.float32)
>>> y = np.array([7, 8, 9, 10, 11, 12]).astype(np.float32)
>>> F.mean_absolute_error(x, y)
variable(6.)

2D array example:

In this example, there are 4 elements, and thus 4 errors >>> x = np.array([[1, 2], [3, 4]]).astype(np.float32) >>> y = np.array([[8, 8], [8, 8]]).astype(np.float32) >>> F.mean_absolute_error(x, y) variable(5.5)

3D array example:

In this example, there are 8 elements, and thus 8 errors >>> x = np.reshape(np.array([1, 2, 3, 4, 5, 6, 7, 8]), (2, 2, 2)) >>> y = np.reshape(np.array([8, 8, 8, 8, 8, 8, 8, 8]), (2, 2, 2)) >>> x = x.astype(np.float32) >>> y = y.astype(np.float32) >>> F.mean_absolute_error(x, y) variable(3.5)