The following node is available in the Open Source KNIME predictive analytics and data mining platform version 2.7.1. Discover over 1000 other nodes, as well as enterprise functionality at
http://knime.com.
Feature Inserter
Inserts features from a data table into the network.
The object to attach the feature to is detected by the available
ids per feature.
- No id -> feature is attached to the graph
- Node id -> feature is attached to node
- Edge id -> feature is attached to edge
- Edge and node id -> feature is attached to the end specified
by the node and edge id
Special features are:
-
label (string): Node/edge feature that specifies the label.
The label is used for the name of an object.
-
is_target (boolean): End feature that specifies the target
node of an directed edge.
-
relevance (float): Node/edge feature that specifies the
relevance of the object [0-1].
Used in the filter nodes to filter objects by setting
the relevance feature to 0 and filtering all objects
for this feature.
-
weight (float): Edge/end feature that specifies the weight
of an edge or node edge pair.
Features should not start with the prefix "knime_"
since this is reserved and used for KNIME internal features
such as the unique network id.
The input table is optional. A single graph feature can be added
without an input table by providing the feature name, type and
value in the node dialog.
An example table that inserts values for nodes, edges and end objects
and that has a column per id, attribute name, type
and value would look as follows:
NodeId |
EdgeId |
Feature name |
Feature type |
Feature value |
node1 |
? |
label |
string |
node1Label1 |
node1 |
? |
nodeFeature1 |
float |
0.1234567 |
node1 |
? |
nodeFeature2 |
integer |
123456 |
? |
edge1 |
label |
string |
edge1Label |
node1 |
edge1 |
weight |
float |
0.5 |
node1 |
edge1 |
endFeature1 |
boolean |
true |
node1 |
edge1 |
endFeature2 |
uri |
www.Network.eu |
In order to use this table select appropriate columns for the
node and edge id, the feature name, type and value.
An example table that contains only one value per object would look
as follows:
NodeId |
Value |
node1 |
0.1 |
node2 |
0.2 |
node3 |
0.3 |
In order to add these values to the nodes select the node id and
value column and enter the name of the feature in the attribute
Dialog Options
- Node ID column
-
The name of the column that contains the node id column.
- The edge ID column
-
The name of the column that contains the edge id column.
- Name column
-
The name of the column that contains the attribute name. If the
none option is selected the user can specify the name in the
Feature name field.
- Type column
-
The name of the column that contains the type information. If the
none option is selected the type for all features is specified
by the Feature type select box.
- Feature name
-
The name of the feature. Only enabled if the none option is
selected in the name column select box.
- Feature type
-
The type for all features. The node determines the feature type
from the data type of the value column if the
'[based on value column]' option is selected.
This option is only enabled if the none option is
selected in the type column select box.
- Value column
-
The column that contains the feature value. Select the none option
if the default value should be used for all objects.
- Default value
-
The default value to use if no value column is selected or if
the value cell is a missing cell.
Ports
Input Ports
0 |
Existing network |
1 |
Optional data table with the features |
This node is contained in Network Base Plug-in
provided by Chair for Bioinformatics and Information Mining,University of Konstanz,Germany.