The Global Stat Parameters page contains all the global parameters for your Statistical Forecasting process, grouped by functional area. This includes the configuration of key transactional data that will influence processes within Stat Forecasting, and the configuration of the customer logo.
System Administrator manage parameters are at a global level on this page. These parameters apply to all product-customer combinations.
Parameters
Effective Parameters are values that are currently set. Edit Parameters allows you to enter new values for the same parameters. When you enter a value, it becomes an effective parameter.
Additional information
- Initialization periods specify how many periods of history to be used to generate the initial forecast.
- If you check Optimize, then the system-optimized values for Alpha and Beta are used at the leaf level.
- If you don't select Optimize for all forecasts at aggregate levels, then the values you entered into Alpha and Beta are used.
- If you choose to dampen the trend for DES, then a manual value for Phi is required.
- Max Annual Growth (%) limits the maximum year-over year growth used in the Calculated % over PY forecast.
- You can override Optimize, Alpha, Beta, DES — Dampen, and Phi values at the calculation level.
2000: Global Parameters
These parameters enable you to specify the history and forecast horizons that display within the tables and charts in the application.
Parameters
| Parameters | Values |
| View Years History | Specify the number of years history to view. If a user enters more than the maximum history available, then the maximum available is shown. |
| View Years Future | Specify the number of future years to view. If a user enters more than the maximum available, then the maximum available is shown. |
| View Period Type |
|
2105: C-VAR Analysis Parameters
The C-VAR analysis parameters enable you to specify the calculation method to be used for C-VAR reporting.
Parameters
| Parameters | Values |
| C-VAR Reporting | Specify the C-VAR measure to be used in D2105 – C-VAR Analysis from these options:
|
| MAPE/Accuracy | Specify the measure for forecast variance:
|
2110: Seasonality Parameters
These parameters enable you to define a planning period to substitute a 53rd week, and apply these forecast methods:
- Trend Seasonal: Calculated percentage over prior year.
- Trend Seasonal: Multiplicative Decomposition.
- Trend Seasonal: Ensemble Trend Seasonal.
- Basic: Prior Year.
2200: Stat Optimization Parameters
These parameters relate to D2200 Stat Optimization. The statistical forecasting engine chooses the optimal Alpha and Beta parameters from a predefined list of options.
Once you select the system best-fit statistical method, you can override the selection.
Parameters
Simple Exponential Smoothing
| Parameters | Values |
| SES – Initialization Periods | Specify a value for the initialization period required to use the Simple exponential smoothing method. Initialization periods specifies how many periods of history to be used to generate the initial forecast. |
| SES – Optimize | If SES - Optimize is checked then the system-optimized values for Alpha are used as the leaf level. If SES - Optimize isn't checked or for all forecasts at aggregate levels, then the values entered into SES - Alpha is used. |
| SES – Alpha | Specify a value as core forecast parameters for the SES forecasting method. Alpha represents the base value which determines the weighting of past data values in setting the baseline for the forecast with higher Alpha values leading to the increased weight being given to the most recent observations. |
Crostons
| Parameters | Values |
| Crostons – Initialization periods | Specify a value for the initialization period required to use the Crostons method. Initialization periods specifies how many periods of history are to be used to generate the initial forecast. |
| Crostons – Optimize | If Crostons - Optimize is checked then the system optimized values for ‘Alpha’ and ‘Beta’ are used as the leaf level. If Crostons-Optimize isn't checked or for all forecasts at aggregate levels, then the values entered into Alpha and Beta are used. |
| Crostons – Alpha | Specify a value as core forecast parameters for the Crostons forecasting method. Alpha represents the base value which determines the weighting of past data values in setting the baseline for the forecast with higher Alpha values leading to the increased weight being given to the most recent observations. |
| Crostons – Beta | Specify a value as core forecast parameters for the Crostons forecasting method. Beta represents the trend value which determines the degree to which recent data trends should be valued compared to older trends when making the forecast. |
| Modified Crostons – Initialization Periods | Specify a value for the initialization period required to use the Modified Crostons method. Initialization periods specifies how many periods of history are to be used to generate the initial forecast. |
| MOD Crostons – Optimize | If MOD Crostons - Optimize is checked then the system optimized values for Alpha and Beta will be used as the leaf level. If MOD Crostons-Optimize isn't checked or for all forecasts at aggregate levels, then the values entered into Alpha and Beta are used. |
| MOD Crostons – Alpha | Specify a value as core forecast parameters for the Modified Crostons forecasting method. Alpha represents the base value which determines the weighting of past data values in setting the baseline for the forecast with higher Alpha values leading to the increased weight being given to the most recent observations. |
| MOD Crostons – Beta | Specify a value as core forecast parameters for the Modified Crostons forecasting method. ‘Beta’ represents the trend value which determines the degree to which recent data trends should be valued compared to older trends when making the forecast. |
Moving Average
MA – Periods: specify a value for the number of periods to use to calculate the moving average when this forecasting method is selected.
Double Moving Average
| Parameters | Values |
| DMA 1 – Periods | Specify a value for the number of periods to use to calculate the moving average when this forecasting method is selected. |
| DMA 2 – Periods | Specify a value for the number of periods to use to calculate the annual moving average when this forecasting method is selected. |
Double Exponential smoothing
| Parameters | Values |
| DES – Initialization periods | Specify a value for the initialization period required to use the double exponential smoothing method. Initialization periods specifies how many periods of history to be used to generate the initial forecast. |
| DES – Optimize | If DES- Optimize is checked then the system-optimized values for Alpha are used as the leaf level. If DES-Optimize isn't checked or for all forecasts at aggregate levels, then the values entered into DES- Alpha are used. |
| DES – Alpha | Enter a value as core forecast parameters for the DES forecasting method. Alpha represents the base value which determines the weighting of past data values in setting the baseline for the forecast with higher Alpha values leading to the increased weight being given to the most recent observations. |
| DES – Beta | Enter a value as core forecast parameters for the DES forecasting method. Beta represents the trend value which determines the degree to which recent data trends should be valued compared to older trends when making the forecast. |
| DES – Dampen | Specify the value for damping factor. Damping factors are used to smooth out the graph and take on a value between 0 and 1. Technically, the damping factor is 1 minus the alpha level (1 – α). |
| DES – PHI | Specify the value for damping co-efficient. This value is only used for advanced data analysis. |
Rolling Line Regression
Rolling regressions are one of the simplest models for analyzing changing relationships among variables over time. They use linear regression but enable the data set used to change over time. In most linear regression models, parameters are assumed to be time-invariant and shouldn't change over time.
RLR Periods: specifies the number of periods to be used.
Calculated % over PY
Max Annual Growth (%): specify a value for a maximum threshold percentage. The system limits you entering values above this threshold relating to max annual growth calculations within the forecasting dashboards.
2400: Best Fit Parameters
These parameters relate to D2410 Best Fit Analysis, that enable the application to consider best-fit forecasting scenarios and ranks them according to the margin of error.
Parameters
| Parameter | Value |
| Forecast Error Basis | The calculation for the forecast error is the absolute forecast error divided by the selected divisor. This parameter specifies the divisor used in the calculation. To change the parameter, select the cell next to Forecast Error Basis and select a parameter from the dropdown:
|
| MAPE/Accuracy | Defines whether to display the Mean Absolute Percentage Error (MAPE) or Forecast Accuracy (defined as 100% minus MAPE). To change this parameter, select the cell next to MAPE / Accuracy and select either MAPE or Accuracy from the dropdown. |
| Lead Time Offset | This parameter specifies the number of planning periods to be used as lead time offset for forecast error calculations. For example: for a forecast made for Week 1, select a Lead Time Offset of 2 to measure the forecast made in Week 1 from Week 3. To change the number, select the cell next to Lead Time Offset and select a number from dropdown. |
| Periods to Sum | This parameter specifies the number of planning periods to sum in the calculation of MAPE. For example: for a forecast calculated in Week 1, select a Lead Time Offset of 2 and select a Periods to Sum of 3 to measure the forecast made in Week 1 offset to Week 3, and then summed for Week 3, Week 4, and Week 5. Note: This parameter references the Initialization Periods set in D2000 - Global Stat Parameters and D2200 - Stat Optimization Parameters. The test window references the initialization period set for the forecast method and discountes the number of periods when producing a forecast. If the Number of Tests is set to 10, and the initialization period is set to 6, then the forecast is based upon the difference between these two values, 4 in this case. |
| Number of Tests | This parameter specifies the number of forecast calculations to be tested. |
2500: Forecast Output Parameters
These parameters relate to the controls used in D2500 Edit Final Forecasts.
Parameters
| Parameters | Values |
| Selected Disaggregated forecast | When you select this, you disaggregate the aggregate best fit forecast with Periods History for Disaggregation. |
| Period history for Disaggregation | When you select this, you can enter the number of history periods for the disaggregation. |
| Disaggregate if C-Var Above Limit | When you select this, you only disaggregate when a C-Var limit is breached. |
| C-Var Measure | You can can select which C-Var measure to use:
|
| C-Var Limit | This enables you to enter the C-Var limit which drives the Disaggregate if C-Var above limit function. |
Add seasonality to Non-Seasonal Method
Overlay seasonality: This enables you to apply seasonality from aggregate to non-seasonal forecasts (based upon enough history).
