About: This article describes how to connect nodes to create process paths in Construct.
Table of Contents
- Introduction
- Connections Methods
- Connection Guidelines
- Disconnection Methods
- Wizard
- Related Articles
Introduction
Nodes need to be connected to one another in order to pass data from node to node through the entire workflow. Some nodes can take connection from multiple other nodes, and some nodes can connect to multiple other nodes. However, the most common workflow consists of nodes with one node connecting in to it and one node connecting from it.
When connecting nodes, the node at the start of the connection line is referred to as the Source Node and the node at the end of the connection line is referred to as the Destination Node. An arrowhead on the connection line indicates the direction data will flow (source to destination). For example in the photo above, the node called Fall2021.csv is the source node and the Cleanse node is the destination node. You can tell because the source node (Fall2021.csv) does not have an arrow pointing to it in the space between the two nodes, while the destination node (cleanse) does have an arrow point to it. In this example the Cleanse node is simultaneously a destination node (in the connection from Fall2021.csv) and a source node (in the connection to Temp Out_3).
Connection Methods
Nodes can be connected several ways:
- Use the Job tab toolbar Connect Nodes tool
- Hold the Alt key+click and drag the source node to the destination node
- Right-click on the source node, select Begin Connecting Nodes, then select the destination node
- Drag and drop a destination node on top of a source node
- Drag and drop a node over an existing connection line between two other nodes. When the connecting line turns bold, drop the node to insert it into the existing path at that point
Important: Make sure the node actually landed on the connection line and is not just "sitting on top" of it. Check this by noticing if each node has a connection line arrow pointing to it.
Connection Guidelines
In order for a job to be configured successfully, it can be helpful to understand a few simple rules for connecting nodes:
- Several nodes such as: Call Job, Command, Script, Zip, and FTP nodes are stand alone nodes and cannot be connected to another node.
- Only a few nodes can accept more than one input. These are: Merge, Append, LookUp, and Report.
Disconnection Methods
There are several ways to disconnect nodes from each other.
- Right-click on a connection line and select Delete from the menu
- Select a connection line and press the DEL key
- Right-click on any node (except those with multiple input or output connections) and select Disconnect from the menu
Note: You can disconnect several nodes at once by highlighting a group and deleting. You can also hold the Ctrl key and select individual nodes, then press the Delete key.
Wizard
Right-clicking on a connection line produces a menu option called Wizard. The Wizard tool can be used to import a scoring model that was built in Rapid Insight's Predict.
Import Scoring Model
When Predict users create a predictive model, it is saved as a ".vpsm” file (Predict Scoring Model). This file contains the mathematical formula and all the related variables that define the model. The model formula can then be applied to a dataset in order to "score" (apply the formula to) the data.
The Import Scoring Model Wizard allows users of Construct to make use of ".vpsm" model files created in Predict. This way, Construct can be used to assign probability scores, then seamlessly be able to report on the results. The wizard translates the contents of the model into nodes – recreating the formula, creating additional variables if necessary, replacing missing data with the appropriate substitute values, etc. so that scoring can take place in Construct.
For more information regarding our predictive modeling and statistical exploration tool, please reach out to RI-Support@eab.com.
Calculate Probability Transform
The wizard tool always creates a Transform node called "Calculate Probability" to incorporate the scoring model formula into the Construct job workflow. If the model formula requires variables that are not present in the data at this point in the workflow, the formula will be displayed in red within the transform node, indicating a problem. Note that this Transform node behaves just like the standard node added by a user. Contents can be edited, removed or the whole node deleted. The scoring results are stored in a new variable called Probability.
Create Variables Transform
If new variables created in Predict entered the model formula, the wizard tool will import an additional Transform node to recreate the same variables in Construct.
Comments
0 comments
Article is closed for comments.