Creating Dataflows - Adding and Organizing Steps

Dataflows are created from steps. Each step processes data in a specific way. Usually, in a multistep dataflow a step further processes objects and properties created or modified in the previous step or steps. The steps are run in order from top to down.

A step in a dataflow can be created either based on a tool or another dataflow. You can combine any number of tools and other dataflows to create a new dataflow.

Learn all about the Dataflow Palette from the dedicated article. Here we go through the basics of adding steps.

Adding a Step Based on a Ready-made Dataflow

Simplebim ships with a growing number of ready-made dataflows for different purposes. You can find them under the Installed Dataflows in the Dataflow Presets dialog. Learn more about each dataflow from the dedicated articles.

Using Tools

Your own custom dataflows can be created by combining the tools. You can find them under the Tools in the Dataflow Presets dialog. Learn more about each tool from the dedicated articles.

Adding a Step Based on Imported or Your Custom Dataflow

When you import a dataflow, it goes to a special folder. For you own dataflows you can create as many folders as you like. You will find them all under the My Dataflows section.

Organizing Steps

In a simple dataflow that only has a few steps you don't have to organize the steps. They are run in order from top to down. 

However, there can be dataflows with hundreds of steps. Managing this kind of dataflow can be difficult if you don't organize it smartly. You have two options here. 

Using Sections

The first one is to use so-called sections in your dataflows. These are like folders, which you use to group steps that relate closely together. 

Usually, you would put all the steps that relate to a specific set of objects, or to a specific operation, under the same segment. For example, defining locations or creating derived objects usually takes multiple steps and creates a meaningful segment to your dataflow.

Segments allow you to run and manage a steps in groups.

Creating Multiple Separate Dataflows 

The other option is to organize your data processing to separate dataflows. There's no hard rule when it makes sense to have one large dataflow and when to manage data processing with smaller dataflows. Simple is better; the dataflow logic can quickly become very complex if you try to handle too many scenarios in one dataflow. 

For example, it might be a good idea to have separate dataflows for each design discipline. One for architectural models, one for structural, and so on... In some cases, the model data is so different even between the model author tools that it makes sense to keep separate dataflows from them. 

You can also group your dataflows in modules based on different kind of operations. Instead of trying to create one dataflow that does everything, you can have a set of smaller modules for different kinds of tasks, which you can then combine as needed.

Finally, even if you manage to create dataflows that are of generic use, there's many times project or even model level adjustments you need to make. This is one more way to organize your dataflows.

How It All Comes Together?

Check also the articles about the ways how you configure the steps and the data flows between the steps.