Running bulk regression analysis on meters

Edit online

The Bulk Regression Modeler - Meters report enables you to perform bulk regression analysis on weather metrics for a selected range of meters, based on a few predefined parameters. The regression analysis in the report works the same way as the meter regression analysis dashboard, with the advantage of running the analysis in bulk.

Before you begin

For each location that is included in the reporting selection criteria, configure the following items:

Link the location to a valid weather station with up-to-date weather data, or sufficient data for the historical period that the regression analysis is run on
Under location settings, configure the location's Base HDD and Base CDD temperatures in Celsius
Under location settings, configure the location's nonworking days if any exist

About this task

For more information about the regression analysis tool, prerequisites, and creating a regression model, see the related link.

Both system administrators and general users of the platform can run the report in Commit To Save mode.

Similar to any regression analysis tool, it is time consuming and computational intensive to perform regression analysis in auto fit mode. In auto fit mode, the tool must try out many combinations of Base HDD and Base CDD values to find a best fit. Generally it is expected that the report times out if the number of meters that are required for regression analysis exceeds 300. Therefore, run the report for a smaller set of data each time, for example, for a Group only rather than for the whole organization. The performance improves substantially if you can predetermine and supply the Base HDD and Base CDD values for each location, rather than using the auto fit method.

You can run the report in two modes, which are View Only mode and Commit To Save mode. In Commit To Save mode, regression results for applicable meters are saved into the platform together with the relevant HDD and CDD metric changes that are required for weather normalization reporting. The report can be output only as an email attachment in CSV format in both modes.

When you run the report in Commit To Save mode, the regression result and the relevant HDD and CDD metric changes are saved into the platform. When you run the report in View Only mode, the meters are marked as To Be Applied.

The following table outlines the items that are saved or altered in the platform:

Entity	Attributes
Meter settings	Slope for HDD, Slope for CDD, Base Load for Working Day, Base Load for Holiday, R2, Regression Results, Regression Base Period
Location Settings	HDD Base Temperature, CDD Base Temperature
HDD and CDD Stats	HDD and CDD stats for the location are recompiled if any meter in the location has its result equal to `Applied`. This is necessary because the calculation of HDD and CDD values relies on the Base HDD and Base CDD values, which might have been changed as part of the exercise.

Note: *The slope or co-efficient values of HDD and CDD are saved in Celsius, assuming that the normalization reporting uses HDD and CDD values in Celsius. Therefore, the value that is saved might be different from the value that is displayed in the report if it is in Fahrenheit. To obtain the corresponding Fahrenheit co-efficient, divide the Celsius value by 1.8.

Attention: The report excludes any meters that are closed at the time of the report end period. For example, if the report is run for one year that ends December 2018, then any meters that were closed in December 2018 or earlier are excluded from the report.

Procedure

To display the Reports grid, click Report > All Reports.
In the Report search field, enter Bulk Regression Modeller - Meters.
To run the report, click the report name in the grid.
In the Bulk Regression Modeler dialog, select your required options:
Group

Run the report for one or all groups.

Location

Run the report for one or all locations.

Data type

Select a data type to run regression analysis on, for example, electricity kWh. Only meters that belong to the selected data type are included in the report.

Filter By # 1 - Execution mode

Select View Only or Commit To Save mode for the report, where the default is View Only.

Filter By #2 - Coverage
You can either let the report determine the best fit base HDD and base CDD temperatures for your locations, or you can preconfigure the base values before you run the report. Also, you can choose to run the regression analysis for meters that have not had a regression model set up before, or to run the analysis for all meters regardless of whether they have any existing regression models:
- Auto Fit - Only create new models
- Auto Fit - Create new or replace existing models
- Use existing location base HDD and base CDD values - Create only new models
- Use existing location base HDD and base CDD values - Create new or replace existing models
Filter By #3 - R2 acceptance criteria
When you run the report in Commit To Save mode, if a meter's R2 regression result value is higher than or equal to the selected accepted R2 value, the model is saved into the platform. When you run the report under View Only mode, the meters are marked as To Be Applied. By default, the accepted R2 value is 0.75:
- R2 >= 0.75
- R2 >= 0.7
- R2 >= 0.65
- R2 >= 0.6
- R2 >= 0.55
- R2 >= 0.5
- No Minimum R2
Note: R² is the coefficient of determination, which is a statistical measure that indicates how closely data fits a model.
Duration

Default is 1 Year

Ending with

Select a calendar month
The report performs regression analysis on weather data for all active meters that are included in the report selection range. The regression analysis result and the reporting outcome are shown in the CSV file report that is sent in the email. For more information about the information that is included in the report, see the related link.