KNIME Documentation

Data types ​

Knime supports multiple data types. This variety reflects in the multiple node port types, in the column types within a data table and in the flow variable types.

Node port types ​

Nodes can have multiple input and output ports of different types. Each node port type looks different. Only ports of the same type can be connected to each other. Here is a list of the most used port types.

Port typeCarriesIconExample node
TableData tableport tableCSV Reader, Excel Reader
DB DataAn SQL query and allows you to preview the resultDB Data PortDB Reader, DB GroupBy
DB SessionThe current DB Session, including data types, connection properties, etc.DB SessionDB Connector, PostgreSQL Connector
Flow VariableFlow variables passed alongside a connectionFlow Variable(red in/out ports on every node)
Generic portCan connect to all kinds of ports, except table and flow variable.Generic PortCapture Workflow Start, Variable Filter
WorkflowA workflow port objectWf PortWorkflow Reader, Workflow Writer
CredentialAuthentication credentials passed to downstream nodesCredential portGoogle Authenticator, Salesforce Connector
Credential (Knime Hub)A Knime Hub CredentialKnimeHub Authenticator, Space Connector
File SystemFor downstream nodes to perform file system operations e.g. read, writeFile SystemLocal File System Connector
Email SessionDefining outgoing mail server and connection propertiesEmail SessionEmail Connector (Labs), Email Sender (Labs)
Filter definitionA filter definition produced by an interactive component viewFilter DefInteractive filter widgets in a component
ImageAn image objectImage PortJavaScript view node, Table View
NormalizePMML document containing normalization parametersNormalizer PortNormalizer (PMML), Denormalizer
OutlierA model of detected outlier boundaries/statisticsOutlier PortNumeric Outliers, Numeric Outliers Apply
Python EnvironmentA reference to a configured Python environmentPython EnvPython Environment Propagation
ReportReport contentReportReport Template Creator, Report Concatenate

Column types ​

Knime tables support columns of multiple data types. The main column types available are:

  • Number (Integer)
  • Number (Long Integer)
  • Number (Float)
  • String
  • Boolean
  • Collection (List) and Collection (Set)
  • HTML
  • JSON
  • XML
  • URI
  • Path
  • Document
  • Date
  • Time
  • Date&time (Local) and Date&time (Zoned)
  • Duration (Time-based) and Duration (Date-based)
  • Bit Vector and Bit Vector (Sparse)
  • Byte Vector
  • Image (PNG) and Image (SVG)
  • Model (PMML)

Flow variable types ​

  • BooleanType
  • DoubleType
  • IntType
  • LongType
  • StringType
  • CredentialsType
  • FSLocationVariableType
  • Array counterparts of the above: BooleanArrayType, IntArrayType, LongArrayType, DoubleArrayType, StringArrayType

Note that not all column types have a corresponding flow variable type. When converting a table cell to flow variable, if the corresponding flow variable type is not available, the value is casted to StringType.