class InferenceEngine::Data

Overview

This class represents the main Data representation node. More…

#include <ie_data.h>

class Data
{
public:
    // construction

    Data(const std::string& name, Precision _precision, Layout layout = NCHW);
    Data(const std::string& name, const TensorDesc& desc);
    Data(const Data& data);

    // methods

    Data& operator = (const Data& data);
    bool isInitialized() const;
    void setDims(const SizeVector& a_dims);
    void setLayout(Layout layout);
    void reshape(const SizeVector& dims, Layout layout);
    void reshape(const std::initializer_list<size_t>& dims, Layout layout);
    Layout getLayout() const;
    const TensorDesc& getTensorDesc() const;
    const Precision& getPrecision() const;
    void setPrecision(const Precision& precision);
    const SizeVector& getDims() const;
    const std::string& getName() const;
    void setName(const std::string& newName);
    const UserValue& getUserObject() const;
};

Detailed Documentation

This class represents the main Data representation node.

Deprecated The Inference Engine API is deprecated and will be removed in the 2024.0 release. For instructions on transitioning to the new API, please refer to https://docs.openvino.ai/latest/openvino_2_0_transition_guide.html

The NN graphs are di-graphs consisting of data nodes and layer nodes.

Construction

Data(const std::string& name, Precision _precision, Layout layout = NCHW)

An empty constructor (dimensionless)

Parameters:

name	Name of the data node
_precision	Precision of the data
layout	Data layout

Data(const std::string& name, const TensorDesc& desc)

A constructor with tensor descriptor.

Parameters:

name	Name of the data node
desc	Tensor descriptor

Data(const Data& data)

A copy constructor.

Parameters:

data	A data object to copy from

Methods

Data& operator = (const Data& data)

An assignment operator.

Parameters:

data	A data object to copy from

Returns:

An assigned object

bool isInitialized() const

Checks if the current node is resolved.

Returns:

true if resolved, false otherwise.

void setDims(const SizeVector& a_dims)

Sets the data dimensions.

After the current node is marked as resolved.

Parameters:

a_dims

Tensor dimensions to set

void setLayout(Layout layout)

Sets the layout value for this Data instance.

Parameters:

layout

Layout value to set

void reshape(const SizeVector& dims, Layout layout)

changes dims and layout at same time

Parameters:

dims	new dimensions
layout	new layout

void reshape(const std::initializer_list<size_t>& dims, Layout layout)

changes dims and layout at same time

Deprecated Use InferenceEngine::Data::reshape(const SizeVector&, Layout)

Parameters:

dims	new dimensions
layout	new layout

Layout getLayout() const

Gets the layout value for this Data instance.

Returns:

Layout

const TensorDesc& getTensorDesc() const

Gets Tensor descriptor reference.

Returns:

reference to TensorDesc

const Precision& getPrecision() const

Gets a precision type of this Data instance.

Returns:

Precision type

void setPrecision(const Precision& precision)

Sets a precision type of this Data instance.

Parameters:

precision

Precision of the data

const SizeVector& getDims() const

Returns:

data dimensions

const std::string& getName() const

Returns:

name of the data object

void setName(const std::string& newName)

Sets a name the Data object.

Parameters:

newName

Name of the data node

const UserValue& getUserObject() const

Returns:

convenient arbitrary user data holder