17. Examples

17.1. How to use the baseline feature

If a forecast, let’s say the “optimal forecast”, has been submitted to the day-ahead market and we are close to the time, where the intra-day market starts, where we can make adjustments to our day-ahead forecast, it can be helpful to see the total uncertainty of the forecast that has been submitted into the market.

By looking at the uncertainty of the forecast, and the performance of the forecast over the past 6 hours since the forecast has been generated, we can evaluate, whether it is worthwile to make adjustments to the forecast in the intra-day.

The picture below shows such a total uncertainty in relation to the optimal forecast. This means that the optimal forecast has been set as baseline.

17.2. Compose graphs with fixed settings

When using ELFI in the every day routines, it can be time consuming to change time slots, areas, variables etc. Therefore, we have introduced a composer that allows to compose certain graphics and saves the configuration, so that the user can with 2 clicks from the login get to the pre-defined graphs.

To use this feature go to <Tools> and then <Composer>.

17.3. Step-by-step process of composing a graph

The composer provides a step by step functionality, where the user is guided through the setup of creating a specific graph with a numer of selections.

The following is a demonstration on how to create a graph that shows the forecasts for German wind power for the past day and the following 3 days ahead. By following this demonstration, it is easy to understand the concept of it.

Step 1: Select a location

In our example we choose the location “DE”

Click the next button to advance to the next step.

Step 2: Selection of Variable

Here we choose “wind power” out of the list of possible variables.

Click the next button to advance to the next step.

Step 3: Selection of duration

Our target is to create a graph that always shows 1 day in the past and 3 days into the future. To achieve this, we choose a total forecast length of 4 days and a start date that is -1 day. A: move the sliding bar in the first box Forecast duration to 4 days B: move the sliding bar in the second box Starting Point to -1 day

Click the next button to advance to the next step.

Step 4: Selection of Time Zone Here you find a world map that is separated into all available time zones. By slinging over the map, the time zones are highlighted in gray. Select the time zone in which you wish to see the forecast data in.

In our example we shoose Central European Time (CET) as we are in Germany and want to see the data following the German local time.

Click the next button to advance to the next step.

Step 5: Forecast Output Mode

here we can choose, whether we want to save the settings of a graph or whether we want to download the selected forecasts with a 1-click in the future.

For our example, we choose to view the graph later on and therefore click into the squared box inside the box Graph.

Click the next button to advance to the next step.

Step 6: Forecast Output Settings

If you have chosen to create a graph you will now be able to make a few useful selections regarding the way the graph looks like later. The figure shows the selection boxes. The first time you create a graph you will have to go through all of them. Later on, those choices of the graph you have selected last will be set in as default values.

Now just follow the choices that are made for our example to get an idea of which choices are possible.

A Selection of curves to be visible on the graph

The uncertainty information from the Ensemble MSEPS forecasts in form of percentiles are visible by default. You have to now choose which of the lines you wish to see. In our example we choose the min, max, measured and the f1 furve, which is the operational optimal forecast. The curves f2-f4 are pre-defined curves by WEPROG, where f4 often is used for short-term forecasts. However, whether or not there is data behind these additional curves depends on whether they are defined for this loctation.

Click the next button to advance to the next step.

B Selection of the visible grid in the graph

Here you can choose to select no grid, vertical and horizontal grid lines. We choose both vertical and horizontal lines for our graph.

Click the next button to advance to the next step.

C selection of the y-axis mode”

Here you can choose, whether you want to have an automatic adjustment of the y-axis or a fixed value. We choose an automatic adjustment for our example.

Click the next button to advance to the next step.

D selection of the data modes

The data modes selection is only relevant, if you have trained an ANN and want to adjust the uncertainty to the ANN forecast rather than to the raw MSEPS output. The function described in the box provides information about what is done in the adjustment computation.

We select nothing in this case and keep the default.

Click the next button to advance to the next step.

E selection of a Base line

If you wish to see the uncertainty of a forecast in relation of a target value, e.g. the minimum to show the absolute uncertainty, then you can enable the base line function and all curves will be adjusted accordingly.

We choose no base line in our example and keep the default. A description and demonstration of how to use and apply the base line can be found here: base line activation

Click the next button to advance to the next step.

You are now done with the selections regarding the “look” of the graph. Click the next button to advance to the next step.

Step 7: View

You will now see the graph that you have composed. If you are happy with the result, you can click the next button to advance to the last step to save you configuration. If you are not happy, you can go back with the back button and make adjustments to your selection.

Once you are happy with your selections you step to the last page, where you give your graph a label.

Step 8: Store Settings

Now you can give your settings a label, so you can in the future with a 1-click get to the graph at any time.

Once you have saved the label you will be guided to the page, where all you settings are stored and you will find your label in the list.

Once you display a graph you can see the list at the left lower part of you interface.