Left Data Pane

The left pane enables you to select different data sources to work with. The toolbar along the top of the pane provides these actions:

Add or change the data source. If the data source already exists, this action replaces the current data set. For replacement, the data source must match the schema of the original data set that is to be replaced.

For attributes, this action imports an external table with additional attributes for the project. You must assign BY variables before you can add an attributes table. For more information, see Working with Attributes.

For eventsan incident that disrupts the normal flow of any process that generates the time series. Examples of events are holidays, retail promotions, and natural disasters., this action supplements the predefined events. For more information, see Importing Custom Events.

After any data is changed, any pipelines that have been run become Out-of-date and need to be run again.

Check the data source for updates. You can also perform this action by right-clicking the data source and selecting Refresh all data.

Open the selected data source in SAS Visual Analytics for further investigation. This action is not available for predefined events.

You can also perform this action by right-clicking the data source and selecting Explore and visualize.

You can work with these types of data in the left pane:

Source data

The source data for any project must be one of these types.

Time Series

Time series are the most common input data source for forecasting projects. Each record has a date or date-time stamp that is used to assess the historical record of the data and to create forecasts. The source table name is listed under Time series in the left Data pane. Selecting the table time seriesan aggregation of transactional data into specified time intervals and sorted according to unique combinations of the default attributes (BY variables) displays the variables from the table in the middle pane. For more information, see Understanding Time Series Data.

External Forecasts

Sometimes you need to work only with the output data from another forecasting project. For example, you can generate forecasts for a time series data set and then export the data. Using the output data, you can create multiple projects, experiment with different pipelines and modeling nodes, and work with overrides in the forecasta numerical prediction of a future value for a specified time period for each unique combination of BY variable values horizonthe number of intervals into the future, beyond a base date, for which analyses and predictions are made..

The source table name is listed under External forecasts in the left Data pane. Selecting the table time series displays the variables and data from the table in the middle pane.

For more information, see Working with External Forecast Projects.

Attributes

Attributes provide a way for you to visualize and work with subsets of project data based on specific attribute values, or filtersa set of specified criteria that are applied to data in order to identify the subset of data for a subsequent operation, such as continued processing. , that you define. Each project can include these types of attributes.

Project attributes

Project attributes are based on variables from your project data. There are two types of project attributes:

• Default attributes

  BY variables are the initial, default attributes for the project. After assigning the BY variables, they are listed when you select the Default attributes. For more information, see Assigning the Default Attributes.

  If you import additional attributes to your project, the default attributes are merged with the imported attributes table and the Default attributes selection is removed from the left pane.

• Imported attributes

  If you import attributes from another data source, the source table name is listed under Attributes in the left pane of the Data tab. Selecting the source table displays the variables from the table in the middle pane along with the default attributes.

  To add more attributes, see Working with Attributes.

Derived attributes

SAS Visual Forecasting generates attributes for each project. The attributes listed are included in all forecasting projects.

• Descriptive statistics

  SAS Visual Forecasting generates descriptive statistics for the dependent variable as determined by the value of the forecast horizon. You can work with filters based on time series that meet specific values of these statistics.

  For detailed information, see Descriptive Statistics (OUTSUM).

• Demand classification attributes

  These attributes are defined by patterns in the time series that can help improve forecast accuracy. Demand classification uses time series information, hierarchical information, and configuration information as input and generates the output with additional segmentation and series statistics information. For a complete description of these attributes and how they are generated, see Demand Classification.

• Model attributes

  These attributes provide detailed information about the selected forecast model for each unique combination of BY variables. This includes which model was selected for the time series and whether any seasonalitya regular change in time series data values that occurs at the same point in each time cycle., trends, or other patterns were detected. For more information, see Model Attributes (OUTMODELINFO).

• Forecast attributes

  You can select filters of time series in the Forecast Viewer based on forecast attributes. These attributes are generated by each modeling node in the pipeline and are not available in the Time Series Viewer.

  For more information, see Statistics of Fit (OUTSTAT).

Events

An event is an incident that disrupts the normal flow of a process that generates the time series. Examples of events are holidays, retail promotions, and natural disasters. You can create events from a list of predefined events and you can import events from an external data source with custom event definitions. Predefined events are available by selecting Predefined events under Events. For more information, see Adding Events to Your Project.

Models

This item shows the models in the model repository. The model repository includes the following types of models:

• Predefined models: These are models provided by SAS Visual Forecasting.
• Custom models: These models are created by forecasters using SAS Visual Forecasting.

All of the models in the repository can be used in any Interactive Modeling node in any pipeline in one project. For more information, see Using the Model Repository.