Data Manager

Use the Data Manager to save model runs for use later on, to reorder runs, to load parameters from earlier runs and to load runs from external data sources.

To open this dialog box, choose Data Manager from the Model menu, or click on the button on the Run toolbar. You can also click on Manage Comparative Data in the panel for comparative graphs and tables.

Note: If the model hasn't been run, there will be no action that can be taken in the Data Manager other than importing a run from an external source.

Data Strategy

The data strategy you select determines how many runs will be kept before they roll off of the list and what will happen to comparative graphs and tables when runs do roll of the list.

Keep all the variables for the ___ most recent runs – Tells the software how many runs you want to keep in place. Any runs above this number that have not been explicitly saved will be dropped to make way for the new run.

Note All variables will be kept only if Keep all variable results has been checked in the Model Settings Properties Panel

Delete runs over the threshold..., if checked, will cause any runs beyond the threshold specified above to be completely deleted. If this is not checked, and you have any comparative graphs or tables, these runs will only be partially deleted, and the values from them will continue to be displayed in the comparative graphs and tables. Unchecking this means that comparative graphs and table will simply keep adding results until you restore results. Checking it, as it is now, will drop runs past the threshold.

Note When you perform a sensitivity run the software will act as if you unchecked the above checkbox for the duration of the sensitivity run, then change back to the checked behavior for the next run. You might for example normally see 2 runs, then see 4 after running sensitivity, then see two again after a normal run.

Runs displayed in comparative graphs and tables

This is a list of runs that will be displayed in any comparative graphs or tables including the Results Panel

and Causal Lens™. You can reorder this list (except as noted below) by dragging the row to a different position. You can also drag a row to the list on the right. For row in the list there are a number of actions you can take by clicking or double clicking at the appropriate location.

• Delete will delete the run. This action can't be undone. If there are any graphs and tables specifically referencing this run the run will move to the bottom of the list under Preserved/Locked Runs, otherwise it will simply be removed.
• Name displays the name of the run. By default, runs are simply names Run 1, Run 2 and so on. Double click on this entry to change the name. There are no restrictions on the form of the name, but the names must be unique.

• Note If you enter in a name that is already in use the name will not be changed.

• Date Run displays the date and time at which the run was created. You can widen the window to display all of this or hover over it to have it displayed in a tooltip.
• Save, if checked, marks the run as saved. Saved runs are not removed when you restore outputs or do a restore all devices. If you drag a run to the right list it will automatically be saved.
• Status is used to indicate what content remains in the run.
  • Full means the run is complete, with all variables. As long as you have set up the model to save all results you can view everything in this run, and use this run to set model parameters.
  • Part means that only variables appearing in comparative tables and graphs are retained in the run. These graphs will still display values from these runs, though if you add a new graph for a different variable there may not be anything retained to show.
  • Extern indicates that the run was imported from an external source. See the discussion of this below.
  • -- indicates that the run is being retained because there are graphs or tables that specify that run in their lists of data to be displayed. These runs are all listed below the Preserved/Locked Runs line and there is nothing that can be done to runs in this list. They are displayed only for reference.
• Use Parameters is available only for Full runs. Click on this to replace all constant and graphical equations in the model with values from this run. If the model has not changed, then doing this and running again will, with the exception of random functions, give the same results. This is a convenient mechanism for loading in parameter changes you might have made in Stella Live into the model.

Note Flows that are constants may be constrained by the upstream and downstream stocks. When these are restored the largest value the flow took on will be used to set the flow equation. This may not match the original value, but should generate the same results.

Save and hide all runs above moves everything in the list to the list on the right. It is the equivalent of dragging each of them across. Runs below the Preserved/Locked Runs line will not be affected.

Note For graphs and tables that are not comparative, the last run in the list (or the last run before the Preserved/Locked Runs line if that is displayed) will be used. To see the results from a run drag it to that position.

Note If the last run in the list is not Full, then many graphs and tables may simply appear blank.

Saved runs hidden from graphs and tables

This list acts as a holding area for runs that you want to keep but for which you do not currently want to look at results. To see the results in one of these runs drag it to the list on the left.

Saved and hidden runs have the same set of actions as runs in the left list, except that they are saved by definition so there is no save checkbox.

Delete all runs above will delete all the runs in the list and remove them from the list. This action can't be undone. If there are graphs or tables that explicitly reference one of the runs in the list that run will appear in the left window below the Preserved/Locked Runs line.

External Data

Use the Load External Data as a Run button to bring data into a model so you can compare then with simulation results. After clicking on the button you will see a file selection dialog. Choose the files (and sheet if it is an Excel file) you want to import. You can select excel or csv format files.

The format for the data files are the same as the export format described in Setting Up Export Links. The first row or column (depending on the orientation of your data) must be Time or the time units of the model (eg Month), with time value on the columns or rows that follow. The additional rows or columns must contain variable names with values (or blanks) for each time. Arrays must be fully specified.

Note The time axis values should be increasing but do not need to be regularly spaced.

After you select file (and sheet for excel files) the software will attempt to process the data and find the variables that match those in the model. You will receive a warning for each mismatch, and an error if there were no variables found or if there was a problem processing the time column or row.

When successful, the data are loaded into a run with the same name as the file (you can change this name by clicking on it). Any comparatives (again including the Results panel and Causal Lens) will display what was in the imported file.

You can import values from anywhere in this manner. This includes values exported from running other models allowing cross model comparison.

If the imported values are missing values for some time points, they will appear as blanks in tables and be interpolated in graphs (if points only is selected for a graph the missing values will be ignored and points will be drawn only where available).