Network Features

 

Network Features

  Network Features

This page describes the supported feature types and some of the features that have a special meaning such as the edge weight.

Features can be attached to all network elements such as

  • the graph itself,
  • nodes,
  • edges, and
  • end objects (for a description see here).

 

Supported Feature Types

The network plug-in supports the following feature types:

Name Example Default Notes
string Single alphanum text. ' ' The string feature can also contain formatting information such as line breaks etc.
uri 'http://example.com/doc.html#n1' ' ' URI
boolean true, false false  
integer -1, 0, 1 0  
long -1, 0, 1 0  
double 3.14, -23.45, -0.12e15 0.0 Floating point numbers.
Date and Time 31.12.2012   Stores date and time values.
image(png)     Stores png images that can be displayed in the Netwok Viewer node.
image(svg)     Stores svg images that can be displayed in the Netwok Viewer node.
color red, light grey, [r=255, g=255, b=255, a=255] Black Stores color values. Supports x11 color names. The color information can be used in the Netwok Viewer node for coloring nodes, edges and labels.
shape rectangle, circle, triangle, reverse triangle, diamond, cross, asterisk, x-shape, horizontal line and vertical line rectangle Stores shape information. Attached to a node this information is used in the Netwok Viewer node to distinguish different node types.
chemical(SDF)     Stores SDF values that can be displayed as node icons in the Netwok Viewer node.
raw     The raw data is base64 encoded when saved to knet file.
list(basic) [-1, 0, 1], ['Alpha', 'Beta'] [] The list type can be combined with any of the above basic types e.g. list(string)
enum [source, target] [] Enables the allowed values to be restricted for this attribute e.g only male and female are allowed.

Developer can add additional feature types by implementing the FeatureType extension point.

Standard Features

The network plug-in uses a set of standard attributes for special attributes such as labeling and edge weight. The name and meaning of theses attributes are described below:

label (type: string)
A label for the node/edge to be displayed. Returns the ID if no label is explicit specified.
weight (type: float)
The weight indicates the reliability or distance of a connection. It can either be assigned to the edge itself or to an end object, which allows different nodes of the same edge to be assigned with different weights for this edge. The weight is usually treated as reliability within the range [0, 1] with 1 representing the highest reliability.
is_target (type: boolean)
This is an end object feature that marks the target node of an edge. For details see the description of an end object.
relevance (type: float)
The 'relevance' flag is used in the filtering nodes to filter nodes and edges. A relevance value of 0 results in the filtering of the corresponding node or edge.
resource (type: uri)
Can be used to store an objects origin (e.g node or edge).
partition (type: string)
The 'partition' attribute is used to store the partition a node or edge belongs to. The partitions are used to define the type of a node or edge like gene, protein, is_a, ... For a detailed description of the internally used graph representation click here.
node_partitions (type: list(string))
This is a graph feature that contains the name of all node partitions. For a description of the internally used graph representation, click here.
edge_partitions (type: list(string))
This is a graph feature that contains the name of all edge partitions. For a description of the internally used graph representation, click here.
knime_uid (type: list(string))
This is a graph feature that stores the user-defined ID of the network.
knime_uri (type: list(string))
This is a graph feature that stores the user defined uri of the network.
knime_uuid (type: list(string))
This is a graph feature that stores the generated UUID of the network, which is internally used to identify networks within KNIME.
LinkedInTwitterShare

What are you looking for?