Ventana software support forum

If you don't need Pro, you could just uninstall it. Then the OS can''t have any ambiguity about which version to use.

Must be.

I would recommend not including a .vdf in the package, unless you need a base run or dataset for a particular reason. That will ensure that any new run is loaded as the first active dataset. That's the only thing I can think of.

Post your .vpm and we can take a look.

I think you want something like this: Hist GDP = GET XLS DATA( ... ) (or any other data source) ~ $/year Last GDP Yr = INITIAL( GET DATA LAST TIME( Hist GDP ) ) ~ year Future GDP = Hist GDP*EXP( Future Growth Rt*MAX(0,Time-Last GDP Yr) ) ~ $/yr This works because Hist GDP will be held constant after...

There are also more advanced versions - see http://vensim.com/comparison/

You need to create a variable representing the future GDP projection, and switch from the data to the projection after the last data point.

You shouldn't have to do anything special. Custom graphs show the first loaded dataset (unless you specify otherwise), so if you include an irrelevant or broken dataset in the package, that will cause problems. I can't think of another reason. Did you try the .vpm I uploaded?

Strange. This works fine for me, whether I publish from PLE or DSS. What OS are you using? Are you sure that the first loaded dataset is a run from the same model?

Providing web services using the Vensim .dll requires a separate license; we'd be happy to generate a quote (probably easiest to contact us at vensim@vensim.com). Forio provides a hosted online simulation service. Unless you need something pretty unique, my guess is that you could develop something ...

This works fine for me. What versions are you running?

Saving the graphs as a vgd should work, but (a) only for custom graphs, and (b) shouldn't be necessary if the graphs are already embedded in the model.

Can you post a sample?

I think we have a solution - I'll message you as soon as it's ready.

I don't think QUEUE functions will work if orders can be filled in any arbitrary order.

This sounds like a discrete event simulation problem. Another tool might be more appropriate.

In Ventity, you could do this by creating and deleting Backlog entities that store the order properties.

Win7, 8 or 10? (XP won't work.)

One possibility: it could be a conflict with an older version. Doing an uninstall of all Ventity versions in the Control Panel, then reinstalling might work.

We'll look at it right away. Haven't seen this before.

Are these custom graphs or analysis tool graphs of workbench variables? In the first case, you need to be sure that the graphs are embedded in the model, or in a .vgd in the published version. In the second, be sure to include the toolset associated with your environment - if the names don't match, ...

Download a copy at the new Ventity site, http://ventity.biz.

For many real-world processes, the SMOOTH function is a good model, and it's far simpler than anything else.

If you're using this a lot, you can put it in a :MACRO:.

Something like: x = ... ~ stuff cum x = INTEG( x, 0 ) ~ stuff*time mean x = ZIDZ( cum x, time-INITIAL TIME) ~ x The above will differ slightly from the stats tool, due to the one-step lag in integration. To get an exact match, you could add the last time step's value to the accumulation, like: mean ...

You can accumulate the variable in a stock and divide by the elapsed time to get the mean.

You can have both if you define a custom graph, and you can display both on the screen at one time. There just isn't a way to generate two different graphs with one click. If you have few elements, another option would be to use the strip graph tool with the causes & uses options turned off. The...

Sure - just vary what you select in the Subscript Control: https://www.vensim.com/documentation/in ... ?20165.htm

If you need to make it permanent, create subranges & custom graphs.

As Tony said, you can use DELAY FIXED(x,TIME STEP,...) or DELAY FIXED(x,one period,...) to obtain x(t-1). If you're using Euler integration with TIME STEP=1 unit, you can also use a stock or SMOOTH. For example, dX = (Y-2*X)/TIME STEP X = INTEG(dX,initial X) TIME STEP = 1 expresses what you describe...

In your simple case, you could solve the problem analytically, but I assume you're asking about the general case. There isn't any direct facility to integrate over non-time dimensions, except for the LOOKUP AREA function. However, lookups can't time vary, so that's not helpful either. To integrate a...