The data workspace enables users to import and export data from their project. This workspace is only available when the Legacy Compiler opt-in is disabled in project settings.
Users can currently export all model data to CSV format on a per-scenario basis. Each element will be mapped to a column in the CSV, with one row of data per simulation tick.
Users can find the export button in the side panel of the data workspace.
Users can upload multiple CSV files to their project for use in the model.
Importing data into the model currently requires the experimental Dynamo compiler to be enabled in project settings.
The following concepts are introduced by the new data imports feature:
DataFrames: A DataFrame is a collection of named data columns. At the moment only numeric data is well supported.
DataSets: DataSets are collections of DataFrames. At the moment, projects are permitted a single, default DataSet.
Raw data: The raw, parsed CSV data uploaded to a given DataFrame.
Preview data: The data in a given DataFrame as it will be interpreted by the simulation.
Using Data in the Model
Before data can be used in a model, it must be uploaded to a DataFrame. In order to do this, select a newly created DataFrame from the default DataSet. You will be taken to the DataFrame configuration view. Here you can upload your CSV file, and configure how it's interpreted by the simulation (see "Configuring a DataFrame" for more information).
Once the DataFrame has been configured, columns in the DataFrame may be referenced from the formula or computed actual field of any model element. For example:
"Default Dataset"->"DATAFRAME NAME"->"COLUMN NAME"
Configuring a DataFrame
Changing the Name
Click the DataFrame name to edit the name inline. This is the name by which the DataFrame can be referred to in the model formulae.
In order to allow more flexible file uploads, users can adjust the some of the settings used to parse the CSV file:
Delimiter: Defaults to
,but can be changed to support, for example, tab-separated-values (.tsv) files
Use Headers: Interpret the first row of data as column headers
In the "Raw" view, columns may be selected or unselected. Only selected columns are available for use inside the model, and are interpreted as numeric columns.
The simulation engine must be told how to interpret the uploaded data. The user may specify time settings which determine the timesteps represented by each row of data.
If the time settings configured for the DataFrame do not match the time settings configured for the project, the simulation engine will use the interpolation settings to determine which values to use at timesteps for which no data exists.
Hold last value: (Default) The last value in the dataset will be used until the next alignment between simulation time steps and data timesteps.
Fill Zeros: Any values that do not align exactly with simulation time steps will be filled with zeros.