Create an import action for your forecast results

Use this CSV file as a template to set up your import action. 

To create the import action:

1. Use Excel to create a new CSV file.
2. Use the CSV file to create your column headers.
3. Add column headers for the following data items, exactly as shown below:
   • timestamp
   • item_id
   • forecast_action
     Use this column to map the different sets of predictions from each forecast action. 
   • P1, P2, and P3
     These represent the three prediction quantiles you can apply to your forecast.
4. In the CSV file, add a placeholder entry for the timestamp column, and add an actual entry for the item_id column.
   Keep the other columns empty. See the example below (scroll across to see the full table):
5. In the module, select Data then Import to import the CSV file. 
6. Select Upload New File... and upload the CSV file. 
7. In the Set Default File dropdown, select either Admins Only or Everyone. 

In this example, the dropdown option Everyone is selected for the Default File.

This makes sure the import runs correctly(opens external page). Do not skip this step.

8. Select Next.

9. Then select Yes. 

Mapping

Manually map to: 

• Include a selection of seasonality and trend columns
• To exclude all seasonality and trends

To manually map the source columns to the target columns:

1. Select the Mapping tab.
2. Map the source columns to the target columns.
   • See optional seasonal trends.
3. In the Mapping tab, select Only update imported cells. 

In Forecast results line items: 

4. Select Map items manually. 

5. Map the columns to include in your forecast.

6. Select Map on names or codes.

7. Select the Time tab and select Dates.
Ensure the date format is Y-M-D.

8. Select Run import.
This example shows a completed import action. 

The values shown are empty in this example. We have not run the forecast results yet.