Edit Nominal Domain (Dictionary)
Adds (and/or sorts) possible values given by the 2nd data table to the
domain of equally named columns
of the 1st input table. This allows the configuration of downstream
nodes on an amended domain list,
e.g. a histogram that should show an empty bin for a value that is not
actually present in the data. Also,
the sorting on the domain values can be changed. This might be useful
for instance for many predictor nodes,
which add a new column for each possible value to the prediction output
table, whereby the order of the
columns is determined by the ordering in the domain.
Dialog Options
- If domain value columns are not present in data
-
Determines if the execution should either fail or
the column should be ignored if an
included column does not exist in
the input table (1st).
- If column types do not match
-
Determines if the execution should either fail or
the column should be ignored if
the types of an included column are
not equal.
- Domain values (2nd input) will be inserted
-
The ordering of the values in the domain might be
relevant for
downstream nodes, e.g. predictor nodes, which append new
columns
representing the different possible values. This option
determines
the
ordering in the output, whereby the additional domain
values (2nd input) can be either inserted "Before existing domain
values
(1st input)" or "After existing domain values (1st input)".
- Maximum amount of possbile domain values
-
Sets the upper bound of the number of possible
domain values.
An extreme high amount of
possible domain values
may
influence and even crash the rest of the workflow,
since table
specifications are hold in memory. It is highly recommended
to use the
given
default.
Ports
Input Ports
0 |
Input data |
1 |
Additional domain value data |
Output Ports
0 |
Input data with changed
specification.
|
This node is contained in KNIME Base Nodes
provided by KNIME GmbH, Konstanz, Germany.