Averaging the Weather First or Averaging the Energy Forecast
There is a question that has long vexed me: If we calculate a normal series of weather variables by averaging them first and running them through a…
Where Did Durbin and Watson Go Wrong?
I’m sure you have all been there—the spot where you mention autocorrelation and the Durbin-Watson statistic. I saw this recently in some testimony…
Holidays Are No Vacation for the Load Forecaster
Summer has arrived. While your friends are frolicking at the beach, you are struggling in a darkened cubicle to ensure that the load forecast…
Dates and Serial Numbers: Might As Well Leap
In Excel and MetrixND, dates are actually serial numbers. This is obvious in Excel when you format a cell containing a date as a number. In the…
When is the R² Not the R²?
A recently received support email contains a riddle. “When I build a regression model in Excel and remove the intercept, I get a different R²…
Improving Financial Analysis with AMI Data
Smart meter data is more granular and more timely than monthly billing data. Access to this data supports a paradigm shift in forecasting processes,…
A Few Thoughts on P-Values
I recently read an article about P-Values that led to some discussion with my Itron colleagues, along with some thoughts on the matter. In the…
2019 Annual Energy Forecasting Meeting
Two pre-conference classes. Three conference days. Eighteen industry presentations. Forty-nine companies. Sixty-nine attendees. These are…
Developing Net Load Uncertainty Forecasts to Support System Operations
Today, the deep penetration of renewable generation resources system operators are requiring new tools to place reasonable confidence bands around…
It’s All the Same, Only the Names Will Change
Last year, Itron was contracted by a small utility (about 11,000 customers) to construct a 10-year-ahead load forecast for capacity planning. I was…
Capped vs. Uncapped Degree Days
In energy modeling, we often utilize spline variables to capture the non-linear relationship between consumption and temperature. These variables…
Leveraging a Waterfall Approach to Explain the Load Forecast
Our first brown bag seminar of the year is entitled “Leveraging a Waterfall Approach to Explain the Load Forecast”. This session introduces a…
Odds and Sods
Probabilities and odds express the same information in different ways. In science and economics, people tend to think in terms of probabilities,…
What are Weather Groups?
If you’re a user of Itron’s Automated Demand Forecasting System, MetrixIDR, then chances are you’ve probably seen the words “Weather Groups” at some…
A Simple Approach for Addressing Behind-the-Meter Solar Generation
I was recently asked to account for behind-the-meter (BTM) solar generation in a set of day-ahead hourly electric models. Fundamentally, the problem…
Daylight Saving Time: The Bane of the Load Forecaster
Sunrise and sunset times vary from summer to winter because of earth’s tilt with respect to its orbit around the sun. This difference is magnified at…
Ignore Missing Option
The Sum function in MetrixND seems like a complex way to make adding difficult. In a MetrixND transformation, numbers are added by joining variables…
Using Neural Networks to Build Robust Hourly Models
Our first brown bag seminar of the year is entitled, “Using Neural Networks to Build Robust Hourly Models.” Neural networks are flexible functional…
Calculating an Average Annual Growth Rate in MetrixND
After finishing my forecast, I like to present forecast average annual growth rates. These results are calculated by converting the monthly forecast…
Seasonal Sales? Not a Problem.
There is a client who has a reporting issue problem. Let’s call him Ray. His annual sales are to be split into summer and winter sales, but on a…
Forecast Tests in MetrixND
Out-of-sample tests are a useful tool for seeing how well a model performs with data it hasn’t seen before (i.e., data that weren’t used to estimate…
Creating a Nonlinear Growth Variable
I love straight lines. After all, the fastest route between any two points is a straight line. But in forecasting, going from point A to point B…
The Difference Between AR(1) and Lagged Dependent Terms
Recently, a utilities regulator tried replicating a client’s regression model used for forecasting customer growth. The regulator wished to build the…
Using In-Line Transformation in the Report Object
I’m developing a forecast that requires base, high, and low scenarios. After building the base scenario, the high and low scenarios are…
The First MetrixND Newsletter
I stumbled across the first MetrixND newsletter today. In July 1998, Regional Economic Research (RER) was preparing users for the MetrixND 1.2…