A guide to giving the demo for the ResultsManager? to the Arizona folks.

Pre-demo

• The ResultManager? will ALWAYS create results for the cache directory specified in the ScenarioManager? ([OPUS_DATA_PATH]/[config:cache_directory]). Make sure it exists.

• Run the script at inprocess.demo.make_indicators_example before you demo. This script creates and outputs all the intermediate results for the indicators that are currently exposed in the GUI. This will allow you to click and view results for any of the indicators in the results section of the results manager, regardless of whether you went to the "generate results with..." tab and explicitly created it.

Demo

After loading the project file "opus_gui.projects.eugene_gridcell", first show the basic stucture, then go into more depth into the available actions. On loading the Gui, I would recommend closing the leaf elements in the XML trees so that its not so overwhelming.

Basic structure

Demonstrate the basic structure, showing that there are three sections to the results manager (for now).

1. libraries -- definitions of available indicators
   1. model-based indicator libraries such as gridcell or parcel
   2. shared libraries
   3. personally defined indicators
2. source data -- associations of simulation results and years. Will be paired with indicator definitions from the library in order to generate results
3. results -- indicator results that have been generated and can now be viewed using a variety of visualizations

Actions

Now describe/show some of the available actions exposed by right clicking. The most important actions to show are the "generate results with..." and "view results as..." actions, which are described in more depth at the end of this section.

1. clicking on the "my_indicators" indicator library:
   1. add new indicator -- define a new indicator using the expression language and assigning a name. This is not available yet (waiting on dynamic XML)
2. clicking on an indicator definition in a library:
   1. View documentation -- shows a tab that renders the documentation. This is not available yet.
   2. Generate results with... -- see below
3. clicking on Data_sources node:
   1. add new data source -- define a new data source (result name + years). Not available yet (waiting on dynamic XML)
4. clicking on a source data node in the Data_sources tree:
   1. Generate results with... -- see below
5. clicking on a result
   1. view result as... -- see below

Generate results with

Right clicking on an indicator definition or source data node and selecting "generate results with..." will pop up a tab that allows you run some indicators. There are three combo boxes allowing you to specify the indicator definition, source data, and dataset that, together, are needed to compute an indicator. Note that the combo boxes for the indicator definition and source data are parsed automatically from the relevant parts of the XML.

After specifying these three fields, you can hit generate results. This is live--it will actually generate the results using the new indicator framework api. Note that selecting different "source data" types in the combo box does not really do anything--it will always use Eugene baseline.

View results as

Right clicking on a result node reveals the "view results as..." action. You can select a number of visualizations to view through here. I recommend the matplotlib map and the table. After selecting a visualization type, a new tab will be launched where you can browse the results.

Do NOT try to view a map for non-gridcell results -- you will get an error message, although the Gui will not crash. In the future, we will have to gray out these options.