Getting Started: KNIME

Selecting Nodes

Click any individual node to select it and you'll see it highlighted with a blue border and selection handles (small squares) at its corners. To select multiple nodes, click on empty canvas space and drag to create a selection rectangle. All nodes within or touching this rectangle become selected. Alternatively, hold Ctrl (Windows/Linux) or Cmd (macOS) while clicking individual nodes to add them to your selection one at a time.

Once multiple nodes are selected, you can drag them as a group to reorganize your workflow layout. The connections between nodes automatically adjust to maintain proper links as you move things around.

Understanding Node Anatomy and Structure

Every node in KNIME follows a consistent visual design that communicates its purpose, status, and connectivity at a glance.

The Node Body

Figure 4.2 - Anatomy of a node

Each node appears as a coloured rectangular box. The node's name appears prominently at the top, with the node type (like "Excel Reader" or "Row Filter") displayed just below. The background color provides quick visual categorisation: purple/pink nodes typically handle data input and output operations, green nodes perform data manipulation and transformation, blue nodes create visualisations, and yellow nodes manage workflow control structures.

Status Indicators: Understanding the Traffic Lights

The small circular indicator in the bottom-right corner of each node tells you its current state through a simple colour system:

Figure 4.3 — Table of Different Traffic Light status

Sometimes a node shows green status (executed successfully) but with a small warning triangle overlay. This means the node completed its work but encountered non-critical issues. For example, it might have found null values or format inconsistencies that it handled by skipping rows.

Warnings don't stop execution but signal you should investigate. Check the Console panel for yellow warning messages explaining what happened.

The spinning gear or progress indicator appears while a node actively executes. For nodes processing large datasets, this might persist for several seconds or minutes. The progress indicator gives feedback that the node is working

Every node in a KNIME workflow has small connection points called ports. These ports determine what kind of information a node can receive and what it can pass on to the next step. They are the way data, models, database connections, and other objects move through your workflow.

Understanding node ports is essential, because the connections you create between ports define the entire flow of your analysis. Once you learn how to read these ports, you can tell at a glance what a node expects as input and what it will produce as output.

Input and Output Ports

Ports are the small triangular connection points on the sides of nodes. They work like electrical sockets. You plug connections into them to create your data flow.

• Input ports appear on the left side of nodes and receive data coming in. Most nodes have one input port, but some have multiple inputs for different purposes. For example, a Joiner node (used to merge two tables) has two input ports with one for each table you want to combine.

• Output ports appear on the right side of nodes and send data forward to the next step. After a node processes your data, its output port makes the results available for the next node in the sequence. Some nodes have multiple output ports when they produce different types of results or split data in multiple directions.

The triangular shape of ports points in the direction data flows: input triangles point right (into the node), output triangles point left (out of the node).

Port Shape Conventions: Triangles vs. Squares

You may have noticed that some ports are triangular while others are square. This distinction carries meaning about how those ports handle their data.

• Triangular Ports: Data Streams: Triangular ports represent data that flows like a stream from one node to the next. Black triangular ports carrying data tables are the most common example. When you connect triangles, the entire data table passes from the first node to the second, which processes it and produces output through its own output triangle. Triangular ports are active and flowing. Data continuously moves and transforms from left to right across your canvas.

• Square Ports: Objects and Resources: Square ports represent objects or resources rather than flowing data streams. Database connections (red squares), models (brown squares), reports (blue squares), and images (green squares) are discrete objects that get passed as complete units rather than streamed and transformed. Think of triangular ports as conveyor belts moving items that get processed at each station, while square ports are like passing a complete tool or finished product. The tool or product is self-contained, not continuously modified.

This distinction helps you understand workflows at a glance. Chains of triangular ports show data transformation pipelines, while square ports indicate that complete objects are being passed around.

Port Multiplicity: Single vs. Multiple Connections

• Single Input/Output: Most nodes have one input port and one output port, creating a simple chain where data flows in, gets processed, and flows out. A Row Filter has one black input (receives the table to filter) and one black output (sends the filtered table forward). This represents straightforward, linear transformation.

• Multiple Inputs: Some nodes have multiple input ports because they need to combine or compare different data sources. The Joiner node has two input ports for merging tables. The Concatenate node can have multiple inputs to stack several tables together. When you see multiple input ports, think about what data sources that node needs to do its job. A node comparing budget to actuals needs two inputs: budget data and actual data.

• Multiple Outputs: Some nodes produce multiple outputs. A Row Splitter has two output ports: one for rows that matched your criteria and one for rows that didn't match. This lets you route different subsets of data down different paths in your workflow. Multiple output ports give you branching capability. Data can flow down multiple paths simultaneously, each applying different transformations, then potentially coming back together later through a join or concatenate operation.

• Optional Ports: Some input ports are optional, shown with a dashed triangle outline. The node can execute without these connections, but connecting them enables additional functionality. For example, many writer nodes have an optional red flow variable input at the top. Without it, the node writes to your manually configured file path. With it connected, the node uses a dynamic path, enabling automated file naming like "GL_Report_January_2025.xlsx" without monthly reconfiguration. Optional ports tell you: "This node works without this connection, but you can enhance it by connecting something here." As a beginner, you can typically ignore optional ports until you need their advanced functionality.

Port Colours and What They Mean

KNIME uses several port types to handle specialised data and operations. Understanding these variations helps you recognise what different nodes do and how they fit into more advanced workflows.

Figure 4.4 — Different Port Types

• Data Table Ports (Black Triangle Ports): The node shown with black triangular ports on both sides represents the most common port type in KNIME. These black triangular ports carry data tables with rows and columns, similar to what you see in Excel spreadsheets or database tables. This is the port type you'll use most frequently in financial workflows. When you import general ledger data with Excel Reader, filter transactions with Row Filter, summarise amounts with Group By, or export results with Excel Writer, you're working with black data table ports throughout. The triangular shape points in the direction data flows: input triangles point right (into the node), output triangles point left (out of the node). Nearly every financial workflow consists primarily of nodes connected by black ports, forming a chain of data transformations from raw source data through calculations and filtering to final reports. Understanding black data table ports is fundamental because they handle all your actual financial data as it moves through the workflow.

• Database Connection Ports (Red Square Ports): The node shown with red square ports represents a Database Connection node. Unlike triangular black ports that carry actual data tables, these square red ports carry an active connection to a database system like SQL Server, Oracle, or your company's ERP database. Think of this as the difference between having a phone number (the connection) versus having the actual conversation (the data). The Database Connection node establishes the link to your database but doesn't transfer any data yet. You then plug this connection into other database-aware nodes that use it to read or write data. For finance professionals, this is valuable when your financial data lives in databases rather than Excel files. Instead of exporting transactions to Excel first, you connect directly to the database and bring in only the data you need. This is faster, more secure, and keeps you working with current data.

• Flow Variable Ports (Red Circle Ports): The node displaying red circular ports on both input and output sides represents Flow Variable operations. Flow variables are special parameters that control how your workflow behaves without carrying actual data tables. These ports always appear at the top of nodes, separate from the main data ports on the sides. Flow variables let you store and pass along values like file paths, date ranges, calculation thresholds, or any other settings that need to change dynamically. For example, instead of hard-coding "January 2025" in multiple places throughout your workflow, you could store "January 2025" as a flow variable at the beginning, then reference that variable throughout your workflow. When February arrives, you change the variable in one place and the entire workflow updates automatically. The red circular ports connect nodes that create, modify, or use these variables. While this is an advanced feature you won't need immediately, flow variables are what transform a single-use workflow into a reusable, automated process that runs month after month with minimal manual intervention.

• Tree Ensemble Model Ports (Brown Square Ports) The node with brown square ports represents machine learning model operations, specifically Tree Ensemble Models. These are predictive models that analyse historical patterns to make predictions about future outcomes. In finance, you might use these for forecasting revenue, predicting delinquent accounts receivable, detecting potentially fraudulent transactions, or estimating project costs. The node that trains the model produces a brown output port carrying the trained model. You then connect this to a predictor node that applies the model to new data. While more advanced than basic reporting, these predictive capabilities can add significant value to financial planning and analysis as you grow comfortable with KNIME.

• Report Ports (Blue Square Ports): The node showing blue square ports handles Report objects. KNIME's reporting capabilities let you create formatted, professional reports combining data tables, charts, text, and images. These reports can be exported as PDF, Word documents, or HTML files. Report ports carry the report object as it's being built. You might connect multiple data sources and visualisations into a report assembly node, which combines them into a single formatted document. The blue port output connects to a report writer that generates the final file. For finance professionals, this transforms KNIME from a data processing tool into a complete reporting solution. Your workflow can import data, perform calculations, create charts, assemble everything into a formatted report, and save or email it automatically.

• Image Ports (Green Square Ports): The node with green square ports carries Image objects. These ports transport actual image files or graphics generated by KNIME nodes. When you create a chart or visualisation in KNIME, that visual output flows through green image ports. This is useful when embedding charts in reports, saving multiple visualisations to separate files, or combining several charts into a dashboard. For example, you might create a budget-versus-actuals bar chart, output it through a green image port, and connect that to a report node that includes the chart in your monthly variance report. Image ports give you control over how visualisations flow through your workflow and where they end up, whether embedded in reports, saved as standalone files, or combined into dashboards.

  
  

From our deployments: New users overthink port colours. In 90% of finance workflows, you'll only use black data ports. We typically don't introduce database or model ports until week two of training. Focus on the black triangles first.