Search found 2297 matches

by tomfid
Mon Jul 16, 2018 1:24 am
Forum: Vensim
Topic: Subscript error
Replies: 5
Views: 31

Re: Subscript error

I think the problem is exactly what the error message reports - you have different subscripts on the right and left side of the equation. "ps hydrostation small" vs. "ps hydrostation small group." That only works if you have one mapped to the other.
by tomfid
Fri Jul 13, 2018 3:33 am
Forum: Vensim
Topic: SD models for demand forcasting
Replies: 1
Views: 16

Re: SD models for demand forcasting

We do a ton of that, but the models are usually proprietary. http://www.ventanasystems.com/model_typ ... marketing/

"Demand forecasting" is a bit too broad to answer - demand for what? Over what horizon? For what purpose?

Have you searched the SD conference archives?
by tomfid
Thu Jul 12, 2018 1:08 pm
Forum: Vensim
Topic: sensitivity analysis
Replies: 2
Views: 24

Re: sensitivity analysis

I hadn't either, but I assume it refers to this: https://swat.tamu.edu/media/114860/usermanual_swatcup.pdf I'd probably have to go back to the source papers to find out what's really going on, but this discusses parameter ranges as if the problem were univariate. A key benefit of MCMC is that it gen...
by tomfid
Wed Jul 11, 2018 3:34 pm
Forum: Vensim
Topic: confidence bounds with MCMC
Replies: 7
Views: 47

Re: confidence bounds with MCMC

When the residuals are autocorrelated, the likelihood is wrong, because it assumes independence. This biases your confidence bounds; they'll be too narrow because it seems that you have more data than you really do. Global temperature is a good example - there are about 150 years of data, but there ...
by tomfid
Wed Jul 11, 2018 1:00 pm
Forum: Vensim
Topic: confidence bounds with MCMC
Replies: 7
Views: 47

Re: confidence bounds with MCMC

1 - right
2 - https://en.wikipedia.org/wiki/Durbin%E2 ... _statistic
3 - there are other, less complete ways, but nothing built in (this is something we're thinking about)
by tomfid
Tue Jul 10, 2018 9:28 pm
Forum: Vensim
Topic: confidence bounds with MCMC
Replies: 7
Views: 47

Re: confidence bounds with MCMC

MCMC always avoids any restrictions on the joint likelihood of the parameters, i.e. no ellipsoids. However, you can run into numerical problems if the likely volume is a low-dimensional manifold in a high-dimensional parameter space. This is not as rare as we'd hope. (Low acceptance rate could be a ...
by tomfid
Tue Jul 10, 2018 2:41 pm
Forum: Vensim
Topic: Model Calibration
Replies: 10
Views: 81

Re: Model Calibration

sumUtility.mdl
(1.75 KiB) Downloaded 2 times
Sum is 0 within numerical precision tolerances.
by tomfid
Mon Jul 09, 2018 5:54 pm
Forum: Vensim
Topic: Model Calibration
Replies: 10
Views: 81

Re: Model Calibration

This is just the same trick as subtracting the sample mean from a set of data points. The mean of the resulting adjusted data is 0.
by tomfid
Mon Jul 09, 2018 3:32 pm
Forum: Vensim
Topic: Model Calibration
Replies: 10
Views: 81

Re: Model Calibration

Try it!
by tomfid
Mon Jul 09, 2018 2:41 pm
Forum: Vensim
Topic: fluctuated graphs
Replies: 11
Views: 60

Re: fluctuated graphs

No, but you can use VECTOR LOOKUP on an array of constants.
by tomfid
Fri Jul 06, 2018 4:12 pm
Forum: Vensim
Topic: Model Calibration
Replies: 10
Views: 81

Re: Model Calibration

I just realized that I misread your original - I thought you wanted to enforce sum=1, not sum=0.

So:

utility[x] = 0 ~ opt parameter, with some reasonable bounds like -10 to 10
norm utility[x] = utility[x] - SUM(utility[x!])/ELMCOUNT(x) ~ this enforces the constraint
by tomfid
Fri Jul 06, 2018 3:16 pm
Forum: Vensim
Topic: Model Calibration
Replies: 10
Views: 81

Re: Model Calibration

The basic idea is to add a term to the payoff that penalizes the difference between 1 and SUM(utilities[i!]).

I'd try the normalization approach first.
by tomfid
Fri Jul 06, 2018 2:42 pm
Forum: Vensim
Topic: fluctuated graphs
Replies: 11
Views: 60

Re: fluctuated graphs

We would need the dataset (vdf) to view.
by tomfid
Fri Jul 06, 2018 2:30 pm
Forum: Vensim
Topic: Model Calibration
Replies: 10
Views: 81

Re: Model Calibration

Three options: - Define the 5th parameter to be 1-SUM(other 4) - this will only work in some cases; you could wind up with a negative value. - Optimize a set of weights that get normalized to determine the shares: x : (x1-x5) ~ dimension for your parameter vector weight[x] = 1 ~ optimization paramet...
by tomfid
Fri Jul 06, 2018 12:45 pm
Forum: Vensim
Topic: fluctuated graphs
Replies: 11
Views: 60

Re: fluctuated graphs

I rounded off. But you also have only 71 months.
by tomfid
Fri Jul 06, 2018 3:38 am
Forum: Vensim
Topic: fluctuated graphs
Replies: 11
Views: 60

Re: fluctuated graphs

Correct - what I mean is that the stock is lower at the end because it has accumulated the loss from the outflow.

The flow is 2/3 hm^3/month, for 72 months, and the difference is about 48 hm^3, i.e. 2/3*72.
by tomfid
Thu Jul 05, 2018 10:33 pm
Forum: Vensim
Topic: fluctuated graphs
Replies: 11
Views: 60

Re: fluctuated graphs

The behavior looks OK to me. There's less water in the aquifer at the end of the run with upwelling, due to the cumulative loss from that constant outflow.
by tomfid
Thu Jul 05, 2018 5:00 pm
Forum: Vensim
Topic: fluctuated graphs
Replies: 11
Views: 60

Re: fluctuated graphs

The superficial answer is that the difference must be due to the additional flow. Without a working copy of the model, it's impossible for anyone to say more. Can you post the model?
by tomfid
Wed Jul 04, 2018 2:49 pm
Forum: Vensim
Topic: Recovering Vensim license from a crashed PC
Replies: 2
Views: 26

Re: Recovering Vensim license from a crashed PC

You just need to keep track of your license key. Obviously you should have backups. As a last resort, we can look up your key, but we can't send it to anyone other than the verified owner of record on the license.
by tomfid
Wed Jul 04, 2018 2:46 pm
Forum: Vensim
Topic: order of calculation of subscripts
Replies: 6
Views: 50

Re: order of calculation of subscripts

I think inverting the order of the subscript is the only easy option.

It might be easier to do with VECTOR ELM MAP.
by tomfid
Mon Jul 02, 2018 1:12 pm
Forum: Vensim
Topic: Rounding discrepancy in Vensim export?
Replies: 2
Views: 51

Re: Rounding discrepancy in Vensim export?

Of course, this is only relevant if you're simulating in double precision.

I think Excel uses Decimals instead of floats, which might create false precision.
by tomfid
Sat Jun 30, 2018 5:32 pm
Forum: Vensim
Topic: remove warning in unit check
Replies: 4
Views: 80

Re: remove warning in unit check

You can do something like the following:

labor index = labor lookup table( labor / unit person )

where "unit person" is an unchangeable constant with value 1 and unit of people.
by tomfid
Thu Jun 28, 2018 12:53 pm
Forum: Vensim
Topic: unit error
Replies: 3
Views: 100

Re: unit error

You also have delay time in units of tourists*day, which is not a unit of time.

Generally the internal errors in macros (delimited by #) are hard to diagnose. Better to start with the other errors at the bottom of the list of 8.
by tomfid
Wed Jun 27, 2018 6:18 pm
Forum: Vensim
Topic: GET TIME VALUE for PLE Plus
Replies: 9
Views: 281

Re: GET TIME VALUE for PLE Plus

arrivals = arrival table(time/unit time)
unit time == 1 ~ with unit of whatever your time unit is; this suppresses warnings
arrival table ( { make this a lookup containing your data } )
by tomfid
Wed Jun 20, 2018 7:38 pm
Forum: Vensim
Topic: data type
Replies: 5
Views: 204

Re: data type

The user guide link above explains the data options, but the short answer is, set the Sub-Type of the data variable to "Equation" in the equation editor.

A stock is an accumulation, so I'm not quite sre what you're asking.