The following node is available in the Open Source KNIME predictive analytics and data mining platform version 2.8.0. Discover over 1000 other nodes, as well as enterprise functionality at

CSV Reader

Reads CSV files. Use this node if the workflow is used in a server or batch environment and the input files structure change between different invocations. In particular, this includes a variable number of input columns. Upon executing the node will scan the input file to determine number and types of the columns and output a table with the auto-guessed structure.

Note: If you are only using the desktop version of KNIME and your workflow is not meant to be used in a batch or server environment use the File Reader node. The file reader node has more configuration options and guesses meaningful default settings as it scans the input file during node configuration (in the dialog).

Dialog Options

Column Delimiter
The character string delimiting columns. Use '\t' for tab character.
Row Delimiter
The character string delimiting rows. Use '\n' for new line character.
Quote Char
The quote character.
Comment Char
A character indicating line comments.
Has Column Header
Select this box if the first row contains column name headers.
Has Row Header
Select this box if the first column contains row name headers (no duplicates allowed).
Support Short Lines
Select this box if some rows may be shorter than others (filled with missings).
Skip first lines
If enabled, the specified number of lines are skipped in the input file before the parsing starts.
Limit rows
If enabled, only the specified number of data rows are read. The column header row (if selected) is not taken into account.


Output Ports
0 File being read with number and types of columns guessed automatically.
This node is contained in KNIME Base Nodes provided by KNIME GmbH, Konstanz, Germany.