Configuring the HANA Demand Forecasting Component

You can configure the properties of the HANA Demand Forecasting component.

Prerequisites:

  • Server: HANA system (SPS 9+) and corresponding version of the Unified Demand Forecast Application Function Library (UDF AFL).

  • Client: Predictive Analytics 2.4+ installed.

Configuration Process:

When configuring the HANA Demand Forecasting component, it is mandatory to map information about schema, table and column names from your HANA sales table. You perform the mapping in the Variables section of the component's Properties tabbed page.

You have the option to configure the settings in the remaining sections of the Properties tabbed page, and the configurations settings in the Advanced and General tabbed pages.

To configure the component, take the following steps:
  1. In Expert Analytics, connect to a HANA Data Source. This data source is your sales table.
  2. Navigate to the Predict Room.
  3. In the Predict Room, from the Component List select Time Series - HANA Demand Forecasting. Drag-and-drop the HANA Demand Forecasting component to the analysis editor. Alternatively, double-click the HANA Demand Forecasting component. Click OK.
  4. To open the configuration settings, double-click the HANA Demand Forecasting component. Alternatively, on the component click the Settings icon and, from the context menu, select Configure Settings.
  5. In the Properties panel of the HANA Demand Forecasting dialog box, the Forecast Horizon section enables you to set the prediction period. Set the Start and End dates.
  6. In the Variables section, you map information from your sales table to the component. Configure the following settings:
    1. Product ID: Select the string-only column from the input table that contains the product identifier code, which can be up to 60 characters long.
    2. Location ID: Select the string-only column from the input table that contains the location identifier code, which can be up to 60 characters long.
    3. Transaction Timestamp. Select the column from the input table that contains the transaction timestamp, which must be in date or timestamp format.
    4. Unit Sales. Select the numeric-only column from the input table that contains the unit sales figure.
    5. Revenue: Select the numeric-only column from the input table that contains the revenue figure.
  7. Optionally, in the Holidays section, configure the following settings:
    1. Schema: Select the schema for the input table.
    2. Tables: Select a table from the schema.
    3. Views: Select a view from the schema.
    4. Time Stream ID: Select the string-only column from the input table that contains the time stream identifier code, which can be up to 10 characters long.
    5. Public Holiday Key: Select the string-only column from the input table that contains the public holiday key, which can be up to 3 characters long.
    6. Operational Status: Select the integer-only column from the input table that contains the operational status.
    7. Timestamp: Select the column from the input table that contains the transaction timestamp, which must be in date or timestamp format.
  8. Optionally, in the Locations to Holiday Mapping section, configure the following settings:
    1. Schema: Select the schema that contains the table with information about mapping locations to public holidays.
    2. Tables: Select a table from the schema.
    3. Views: Select a view from the schema.
    4. Location ID: Select the string-only column from the input table that contains the location identifier code, which can be up to 60 characters long.
    5. Holiday ID: Select the string-only column from the input table that contains the holiday identifier code, which can be up to 10 characters long.
  9. Optionally, in the Demand Influencing Factors section, configure the following settings:
    1. Schema: Select the schema that contains the table with information about Demand Influencing Factor..
    2. Tables: Select a table from the schema.
    3. Views: Select a view from the schema.
    4. Product ID: Select the string-only column from the input table that contains the product identifier code.
    5. Location ID: Select the string-only column from the input table that contains the location identifier code.
    6. DIF Tag: Select the string-only column from the input table that contains the demand influencing factor tag, which can be up to 32 characters long.
    7. DIF Attribute: Select the string-only column from the input table that contains the demand influencing factor attribute, which can be up to 32 characters long.
    8. Timestamp From: Select the date-only column from the input table that contains the date that the timestamp begins.
    9. Timestamp To: Select the date-only column from the input table that contains the date that the timestamp ends.
    10. DIF Value: Select the numeric-only column from the input table that contains the Demand Influencing Factor (DIF) value.
  10. Optionally, in the Expected Future Prices section, configure the following settings:
    1. Schema: Select the schema that contains the table with information about Expected Future Prices.
    2. Tables: Select a table from the schema.
    3. Views: Select a view from the schema.
    4. Product ID: Select the string-only column from the input table that contains the product identifier code.
    5. Location ID: Select the string-only column from the input table that contains the location identifier code.
    6. Timestamp From: Select the date-only column from the input table that contains the date that the timestamp begins.
    7. Price: Select the numeric-only column from the input table that contains the price.
  11. Optionally, navigate to the Advanced tabbed page. In the Configuration Parameters section, configure the following settings:
    1. Damping factor (FC_TREND_DAMP): Defines the damping factor for the trend regressor. The range is value >= 0.00000.
    2. Total regress mass (MOD_HDM_NEAR_HOLIDAY_DENSITY): Sets the proportion of total regressor mass to the right (POST: left) of the half-way date. Note that there are two groups of HDM-regressors, SYS:CAL:YR:HDM:PRE:* and SYS:CAL:YR:HDM:POST:*. The PRE-regressors define the ramp-up before the holiday, the POST-regressors the ramp-down after the holiday. The range is 0.50000 <=value < 1.00000.
  12. Optionally, in the Time Delay Effect section, configure the following settings:
    1. Observation weight (MOD_TIME_WEIGHT): Sets the weight of a one-year-old observation in modelling, compared to an observation taken today. This way, the variable helps to decide whether to give equal importance (or weight) to all the records irrespective of their timestamps. For example, when building the model, setting the parameter value to 1 gives equal importance to all the records irrespective of the time that they were recorded. Whereas, setting a value to the parameter less than 1 enables the user to give less importance to the records that have older timestamps, as compared to the records that have recent timestamps. The range is 0.50000 <=value < 1.00000.
    2. Lower Boundary on weight (MOD_TIME_WEIGHT_MIN): Defines a lower boundary below which the weight will not fall. The range is 0.00001 <=value <= 1.00000.
  13. Optionally, in the Out of Stock Detection section, configure the following settings:
    1. Zero sales period(MOD_OOSD_MIN_LEN): Sets the minimum length of the continuous zero sales to be considered for an out-of-stock period evaluation. The range is value >= 1.00000.
    2. Probability threshold (MOD_OOSD_THRSHLD): Sets the threshold for the probability score to determine whether an item is out-of-stock. The probability score for each item is derived based on the occurrence of zero sales for a period greater than the values specified in the parameter, MOD_OOSD_MIN_LEN. The range is value >= 1.00000.
  14. Optionally, select the appropriate Time Series Decomposition checkboxes to decompose and clearly see the influence on your results of Seasonality, User Promotion and Holidays. De-select the checkboxes if you do not want to consider the impact of these factors in your results.
  15. Optionally, in the Outlier Detection section, configure the following settings:
    1. Configure outlier detection: Defines whether to switch on or off outlier detection.
    2. Acceptable distance from mean (MOD_OUTLIER_MEAN_FACTOR): Defines the outlier detection factor to determine how far away from the mean is acceptable.
    3. Min. non-zero observations (MOD_OUTLIER_STD_DEV_FACTOR): Defines the outlier detection minimum number of non-zero observations, regular and promotional, counted before 0-filling.
    4. Standard deviation factor(MOD_OUTLIER_STD_DEV_FACTOR): Defines the outlier detection factor to determine how many deviations away from the mean is acceptable.
  16. When you have configured the necessary settings, click Done.
  17. Optionally, in the General tabbed page, set properties in the Basic section, such as a component Alias Name and Description.
  18. Click the Run Analysis icon.
  19. When the analysis executes, click OK on the notification message.
  20. Click the Results tab to view the results.

View the Results Grid:

In the Results tabbed page, the results grid is shown by default. For a description of each column in the grid, see Results Grid.

View the Algorithm Summary:

Click Summary to see an overview from the algorithm that describes the behavior of the product location combinations in relation to price elasticity. For a detailed description, see Algorithm Summary.

View the Graph

Click Model Representation to see a graph that displays data points for both the historical data and the forecast range. You can zoom-in on the graph to isolate any portion. For example, to zoom-in on the predicated values for the Forecasted Sales (which are graphed in yellow lines at the end of the graph), select the final portion of the Slider control bar under the graph. The graph will change focus to display the Forecasted Sales.

Export the Analysis:

You can export a Demand Forecast analysis as a stored procedure. For a detailed description, see Exporting an Analysis as a Stored Procedure.

You can now configure HANA Demand Forecasting component to forecast future sales.