Delayed System time

[mdsdoha]

Hi!

I have faced one more problem while I am trying to do reality check through comparing real output with model output. Say, I have given input through lookup of using time 1 to time 100. My model is bit long and complex. If I run the model, the outcome of some variable shows straight zero in first 5/6 time step. Then it start pick up the value. The problem is system take 5/6 time step to flow the output value for that variable which I do not want to. Can someone help me to find out a way that whole model will start with TIME STEP 1 and also start producing output at the same time?

thanks

Doha

[Administrator]

You will need to upload the model if you want us to help.

[LAUJJL]

HI

If in the reality the outputs are delayed, you must do the checking after a while like;

Name of the reality check name :the condition: time >= 0 :implies: the condition that must be verified :OR: time <= the time when the checking starts.

Regards.

JJ

[mdsdoha]

Dear Admin

Here, I have attached my model. If you run this model and see the result of later variable such as 'dist1', dist2' and so on, you will find the results are zero for first couple of time step. I think, system is pushing the time as supply goes far. But I want instant effect like giving input in time step 1 and will give the result for all variable at the same time. Is it possible? Please help.

Thanks

You will need to upload the model if you want us to help.

[LAUJJL]

You cannot get information about future values, unless they are easy to calculate at the beginning of the simulation, but in this case your model is no more dynamic but static and you do not need stocks and time steps any more.

JJ

[mdsdoha]

Hi JJ

Thanks. But what else will need to make my model dynamic? Can you give few clues? If I do not require any stock then what else can I use instead of level? I did not get answer relating why number of initial time step results shows as Zero?

Thanks for helping my learning endeavour.

You cannot get information about future values, unless they are easy to calculate at the beginning of the simulation, but in this case your model is no more dynamic but static and you do not need stocks and time steps any more.

JJ

[LAUJJL]

All your levels are initialized to 0, and it is then expected that these levels are equal to 0 at time equal 0.

What are you trying to check in your model relative to a known behaviour of the reality under certain conditions?

One more remark. Reality checks are powerful, but to my opinion it is necessary before trying to use them to have a minimum of knowledge of basic modeling in Vensim for instance.

JJ

[mdsdoha]

Hi JJ

Thanks for your reply. I am a apprentice in modelling and Vensim. Hope all of you will help at least minimum. As you have written that all of my level are 0, even I put any value in initial then it might go one step and again it will shows 2/3 time step zero which is not looks realistic. Could you please give clues based on my model what possible addition will make this model minimally perfect? So far I have read, reality check will comes up in checking model reliability. Is it right?

In addition, you also mentioned that future is not predictable, then my question is how weather forecast is done by simulation?

Thanks again and waiting to hear from you.

All your levels are initialized to 0, and it is then expected that these levels are equal to 0 at time equal 0.

What are you trying to check in your model relative to a known behaviour of the reality under certain conditions?

One more remark. Reality checks are powerful, but to my opinion it is necessary before trying to use them to have a minimum of knowledge of basic modeling in Vensim for instance.

JJ

[LAUJJL]

You should start by tracing your variables, by using the table tool on the left side of the sketch. right click on it and set the selected variables to causes and set print every to 0.

Then select one after the other every avariable and click on the table tool to print their values every time step and the values of the input of its equation and verify by clicking on the document tool at the left side of the sketch too, that the values correspond to the equation.

Doing this will probably give some answers to your question.

Regards.

JJ

[mdsdoha]

There are so many things to learn. Thanks again to open up another learning for me. Could you please explain what is" set print to 0"??

You should start by tracing your variables, by using the table tool on the left side of the sketch. right click on it and set the selected variables to causes and set print every to 0.

Then select one after the other every avariable and click on the table tool to print their values every time step and the values of the input of its equation and verify by clicking on the document tool at the left side of the sketch too, that the values correspond to the equation.

Doing this will probably give some answers to your question.

Regards.

JJ

[Administrator]

" set print to 0"?

If you right click on the table tool, there is a setting for "print every". Set this to be zero.

[mdsdoha]

Hi!

It is Zero, after that the problem remain same.

thanks

" set print to 0"?

If you right click on the table tool, there is a setting for "print every". Set this to be zero.

[LAUJJL]

I do not understand what do you mean by:

'It is Zero, after that the problem remain same'

Are you able to trace the variables as I explained it?

JJ

[mdsdoha]

Hi!

I meant that it has already given 0 by default. What I tried is that I run the model and click each variable and see table values and found print every as 0. was it like that you meant? If no, could you please tell step by step so that I can understand as an apprentice.

Is there any other explanation to get such result? In my observation of simulation result, one rate in , level and rate out push back one zero for following step result. I have attached the model before and you can see the result was like that. This is the way, three step of supply chain are given 3 zero first then start picking the number, fourth supply chain giving four zero and so on.
Sorry for your inconvenience if I failed to make you understand.

I do not understand what do you mean by:

'It is Zero, after that the problem remain same'

Are you able to trace the variables as I explained it?

JJ

[Administrator]

It would really help you a lot if you went through the whole tutorial at
http://www.public.asu.edu/~kirkwood/sysdyn/VenPLE.pdf

You need to look at the numbers and work out why they are not what you expect. The tutorial will take you though building a simple model and examining the output.