cc.redberry.core.tensor

Interface TensorFactory

All Known Implementing Classes:

ArcCos.ArcCosFactory, ArcCot.ArcCotFactory, ArcSin.ArcSinFactory, ArcTan.ArcTanFactory, Cos.CosFactory, Cot.CotFactory, Exp.ExpFactory, ExpressionFactory, Log.LogFactory, PowerFactory, ProductFactory, Sin.SinFactory, SumFactory, Tan.TanFactory

public interface TensorFactory

All tensors in the system are reduced to the general form. For example: if there are several numerical factors in the product they will be multiplied and only one resulting numerical factor will be in the resulting product. To maintain this reduced form of expressions, specific infrastructure for tensor creation is provided. All creations of tensors are performed using this infrastructure and explicit tensor creation is architecturally forbidden for the user. Tensor creation infrastructure consists of two functionally redundant but optimized for different usage strategies interfaces: TensorBuilder and TensorFactory. In fact TensorBuilder and TensorFactory returned by one particular tensor will produce exactly the same result for the same input sub-tensor sequence.

This class determines the common interface for tensor factories. Objects of this type are produced by Tensor.getFactory() method.

Main contract for the factory infrastructure could be expressed in the following code:

 Tensor tensor = ....;
 TensorFactory factory = tensor.getFactory();
 Tensor createdTensor = factory.create(tensor.getRange(0, tensor.size()));
 assert TensorUtils.compare(createdTensor,tensor);

So, using a tensor factory of any tensor you can rebuild it into the equivalent tensor.

Behavior of TensorFactory is fully consistent with TensorBuilder, see TensorBuilder for more information. For general tensor creation use factory methods in Tensors class.

Since:

1.0

See Also:

Tensor.getBuilder(), TensorUtils.equals(cc.redberry.core.tensor.Tensor, cc.redberry.core.tensor.Tensor), http://en.wikipedia.org/wiki/Builder_pattern

Method Summary

Modifier and Type	Method and Description
Tensor	create(Tensor... tensors) Creates tensor with provided array of sub-tensors.

Method Detail

create

Tensor create(Tensor... tensors)

Creates tensor with provided array of sub-tensors.

Parameters:

tensors - array of sub-tensors

Returns:

created tensor