Row Sampler

This node extracts a sample (a bunch of rows) from the input data. The dialog enables you to specify the sample size.

Options

Output size type

Defines how the size of the output is specified: as a percentage of total rows (relative) or as an absolute number of rows.

Relative size

Specifies the percentage of rows from the input table to be included in the output. Must be between 0 and 100 (inclusive).

Number of rows

Specifies the absolute number of rows to include in the output. If the input table contains fewer rows than specified, all rows are placed in the output.

Sampling strategy

Determines how rows are selected for the output. Strategies include random, stratified, linear, and first rows (sequential).

• Random: Randomly selects rows from the input table. You can optionally specify a random seed for reproducible results.
• Stratified: Preserves the distribution of values in the selected group column. You can optionally specify a random seed for reproducible stratified sampling.
• Linear: Selects rows evenly spaced across the input table, always including the first and last row. This method is useful for downsampling sorted columns while preserving boundary values.
• First rows: Allows you to select the top-most rows of the input table.

Group column

Specifies the column whose value distribution should be preserved in stratified sampling. Ensures both selected and non-selected rows reflect the same distribution of values.

Fixed random seed

Optional seed value for random or stratified sampling. Using a seed ensures the same rows are selected each time the node is executed. Without a seed, a different random selection will occur each time.

If input table is empty

Defines how the node should behave when the input table has no rows. Options include generating empty outputs or failing the execution.

Views

This node has no views

Workflows

• _01_ChemicalLibraryEnumerationKNIME Hub
• 00 Data GenerationKNIME Hub
• 00 Data GenerationKNIME Hub
• 00_Guided_Labeling_for_Document_Classification_(legacy)KNIME Hub
• 00_Keras_Transfer_LearningKNIME Hub
• Show all 779 workflows

Links

• Selecting the top 50 rows of a table by a control columnby armingrudd on 2019-11-05

  forum.knime.com/p/74011
• Partitioning absolute number of rows from top slowby izaychik63 on 2021-04-12

  forum.knime.com/p/120671
• Create a random list of valuesby s.roughley on 2019-05-03

  forum.knime.com/p/59889
• How to Filter Databy HansS on 2021-08-28

  forum.knime.com/p/133011
• Unbalanced Databy armingrudd on 2019-10-26

  forum.knime.com/p/73300
• Show all 6 links

Developers