Column Filter

The Column Filter is one of the most commonly used nodes in KNIME. Although simple nature--selecting and deselecting columns--it is invaluable for many workflows. The Column Filter node allows users to select columns in three different ways:

1) Manual Selection: You manually choose which columns to include or exclude. Pay attention to the "Enforce Inclusion" and "Enforce Inclusion" options. If you have dynamic data sets whose columns might change, you may want to enable one of these options to ensure no new columns appear where they should not.

2) Wildcard/Regex Selection: By writing a wildcard expression or regular expression you can control which columns to include or exclude dynamically. In the second example to the left, the wildcard expression "State*" ensures that any time a field matches that pattern that it will be included in the column set. All other columns will be excluded automatically.

3) Type Selection: By indicating which types of columns you wish to include, you can dynamically control the columns that pass through this node. Type Selection is a valuable asset when you want only numeric columns to flow through, as with certain predictive models.