Product:
Planning Analytics 2.1.19
Issue:
How to get a Forecast for a cube view?
Solution:
To make a forecast happen, you need to have a few tings setup in your TM1 model. This is a list of suggestions that may be needed.
The cube you are doing the forecast must have a TIME dimension. This are in TM1 Architect set on the cube properties.

The dimension for time, need to be in a format that TM1 understands.
Like Only years, or Year – quarters – months hierarchy.
Forecasting requires the time dimension to be placed on the column.
You must have data in all the periods, if you only have data in Quarters, but the dimension contain blank months, the predict may not work. You can get a message like:
Too many missing or zero values in the historical data.
The historical data must represent at least double what is being forecasted. If you forecast 2 years, you need 4 years of old data.
The view need empty periods where the prediction can be created. If you want to look 5 years forward, your TIME dimension need to contain empty elements for 5 years.
To begin forecasting, click an exploration view in a book, then click Forecast and click either Univariate or Multivariate, depending on which type of forecast you want to generate. A univariate forecast predicts the future values of a single variable based on historical data, using the Holt-Winters model.
Use the Set up tab to define the start date and end date for your forecast.
https://www.ibm.com/docs/en/planning-analytics/2.0.0?topic=options-set-up-forecast-tab
You can preview only one time series at a time. The forecasting preview shows what the results for the forecast look like, based on your input.
https://www.ibm.com/docs/en/planning-analytics/2.0.0?topic=forecasting-preview
If you do not have a }DimensionAttributes cube in your TM1 model, you need to create one from PAW.
Right click on the Dimension and select Dimension Attributes.
Enter the attribute name: DIMENSION_TYPE

Click Add.
Then in the view of the DimensionAttributes cube, find the row for your year dimension and enter TIME as DIMENSION_TYPE.
This is the TIME dimension that can be used by the forecast.
https://revelwood.com/ibm-planning-analytics-tips-tricks-planning-analytics-forecasting-configuring-the-time-dimension/
You can also use PAW to create a TIME dimension for you – that can be on monthly level.

The TIME dim created by PAW will look like this, you may use it in your new models.

Above is only suggestions, depending on the version of PAW this can behave or look different.
More Information:
https://revelwood.com/use-the-planning-analytics-ai-assistant-for-smarter-data-analysis/
https://www.ibm.com/docs/en/planning-analytics/2.0.0?topic=dimensions-time-in-planning-analytics-workspace
- Right-click the Dimensions node in your database on the Databases tree, then click Create time dimension.
Alternatively, you can create a time dimension by clicking the Configure time dimension option when you create a new dimension. The Configure time dimension option is only available for a dimension with one hierarchy and no members.
- If required, enter a name for the dimension.
- Select the Start year for the dimension.
- Select the End year for the dimension.
- Answer the How frequently do you plan? question by selecting the granularity of the dimension.
- Months
- This option gives you three levels: Months, Quarters, Years.
- Quarters
- This option gives you two levels: Quarters, Years.
- Years
- This option gives you one level: Years.
- Check the preview to confirm that it matches your expectation, then click Save.
You can modify the time dimension in the dimension editor after you create it.
https://www.ibm.com/docs/en/planning-analytics/2.0.0?topic=forecasting-requirements
Forecasting requires the time dimension to be placed on the column. The algorithm requires the time dimension members to be organized in order of increasing time, where each member of a level represents a consistent unit of time. Members within the dimension definition that do not conform to this requirement, for example calculated year-to-date member, can be hidden. Relevant time periods can be chosen to be ignored, resulting in a linear interpolation of values to be predicted from. Forecasting also supports hierarchical time dimensions and nested dimensions as time: year, quarter, and month dimensions.
One consequence of this is that leaves with multiple parents along the time series would result in non-contiguous time, so special care must be taken in these cases. If possible, it is advised to consider adding a special hierarchy for time series forecasting to support the required time configuration.
https://www.ibm.com/docs/en/planning-analytics/2.0.0?topic=views-forecasting
https://www.ibm.com/docs/en/planning-analytics/2.0.0?topic=forecasting-error-messages
- Something went wrong while generating the forecast for one or more series
- This is a general message for any unsuccessful prediction. Please try again.
IBM Planning Analytics Tips & Tricks: Planning Analytics Forecasting – Configuring the Forecast Settings
IBM Planning Analytics Tips & Tricks: Planning Analytics Forecasting – Previewing the Forecast
IBM Planning Analytics Tips & Tricks: Planning Analytics Forecasting – Setting up the Forecast