# Difference between revisions of "Running Your Model"

(→Run Buttons) |
(→Clock Settings) |
||

Line 14: | Line 14: | ||

Nova supports 3 different integration methods | Nova supports 3 different integration methods | ||

+ | <blockquote> | ||

;Euler | ;Euler | ||

:Current state values are used in computations to determine the next state. Used for all discrete models. | :Current state values are used in computations to determine the next state. Used for all discrete models. | ||

Line 20: | Line 21: | ||

;Discrete | ;Discrete | ||

:This is not actually an integration method; rather it sets '''Dt''' to 1 and uses the Euler method. The result is a sequence of State values rather than an attempt to approximate a continuous function. | :This is not actually an integration method; rather it sets '''Dt''' to 1 and uses the Euler method. The result is a sequence of State values rather than an attempt to approximate a continuous function. | ||

+ | </blockquote> | ||

==Run Buttons== | ==Run Buttons== |

## Latest revision as of 22:01, 10 January 2015

This section describes how you run a model built with Nova's visual interface. It also describes several useful run-time features.

## Clock Settings

Once you have built a model but before you can run it as a simulation, you must first set the clock parameters. The clock settings consist of the following 4 parameters:

**Start**The time (in model time) when the simulation starts (this is usually 0).**End**The time (in model time) when the simulation ends.**Dt**The unit of increment (in model time) for updates.**Integration Method**The mathematical algorithm used to update Stocks and Variables.

Nova currently uses an abstract unit of time ("model time") that serves as the equivalent of some model-dependent real time unit (seconds, minutes, hours, days, years).

**Dt**, the increment, or "delta t", value, is the amount of time that elapses between State updates. For discrete models this is usually 1 (see below), but can be any value (0.1, 0.01, 0.05, 2, 20).

Nova supports 3 different integration methods

- Euler
- Current state values are used in computations to determine the next state. Used for all discrete models.
- RK2, RK4
- These
Runge-Kuttaalgorithms treat Stocks and Variables as continuous functions and reduce the error by subdividing the update interval. Simulations run slower but with more accuracy. RK4 is the preferred option of the two.- Discrete
- This is not actually an integration method; rather it sets
Dtto 1 and uses the Euler method. The result is a sequence of State values rather than an attempt to approximate a continuous function.

## Run Buttons

Once you have built a model in Nova, you can run it by clicking **Capture** → **Load** → **Exec.** Let’s break it down and see what each of these buttons does.

- Capture
- This button “captures” the structure from the components of the model you’ve built and turns it into a NovaScript program.
- Load
- Clicking
Loadloads the information captured from the model and prepares it to run.- Exec
- Clicking
Execexecutes the run command on the model. You will be unable to selectExecuntil you have first selectedCaptureandLoad.- Init
- You can also run the model by clicking
Capture→Load→Init.Initinitializes the simulation run. After you’ve selectedInit,you can either selectRunto run the model at the selected speed, or clickStepto run the model one time-step at a time.

When you are in timeline mode, you can use the **Back** button to go back exactly one time step.

## Timeline

The timeline feature allows you to move back and forth in time over the span of the model. To access this feature, click the Timeline button before you’ve started running the model.

By dragging the slider around on the timeline, you can visually and numerically see the output values for your model at specific time steps.

To go back in time, first press stop. Then just click or drag the time marker to an earlier time point. You’ll see the simulation clock play backwards and the simulation return to the selected earlier time point. To go forward, click run, click on a later time point on the time line, or drag the slider to a later time point.

When you run the model in timeline mode, you can use the **Back** button to move back exactly one time step. Click **Capture,** **Load,** **Init,** then **Run** or **Step.** You can then use the **Back** button to go back in the model, one time-step at a time.

- Limitations

The timeline will work with models that have some element of random chance, like the Firestarter or Logistic growth model, but will not duplicate the sequence of random numbers when the model is repeated. Timeline does NOT work with models that have births and deaths.

Here is a video demonstrating the Timeline feature.

## Automode

When AutoMode is selected, you can adjust values in the model, and see the changes instantaneously in the output, without having to click any buttons to rerun the model.

You can select AutoMode after you have run the model or after clicking **Capture.**
Then you can use sliders to adjust parameter values and see instantaneous results.

Here’s a quick video that shows you how this works for the Logistic model in the Nova model library.

## IMode

IMode shows you how the values of individual components change during the running of the model. IMode shows the values of terms and flows numerically. Terms and DataInput pins can be selected to appear as sliders on the model canvas. The values in Stocks are shown as a line graphs.

To use IMode, click **Capture** → **Load** → **IMode** → **Exec**

Click here to view a video demonstrating IMode.

- To change a Term or DataPin component into a slider while in Imode

- Right-click to open the component's properties;
- Check the interactive box