Hi,
I try to use vensim to re-build the model of Future Electronics which was in dynamo under the textbook by Michael goodman Study notes in SD.
But, I face several problems.
a. 2 units errors discovered - I cannot figure out what is wrong as I follow the unit mentioned in the text.
b. the effective Testers become negative - seems something missing. In the text , it does not mention any initial stock qty about the testers.
c. I cannot understand why hiring rate (HR.KL) would have the initial seting as the text mentioned below
HR.KL=(FHM.K)(TT.K+T,K) 1,R
HR=20/69 1.1,N
HR - HIRING RATE (MEN/MONTH)
FHM - FRACTION HIRED PER MONTH (FRACTION/MONTH)
TT - TESTERS IN TRAINING (MEN)
T - TESTERS (MEN)
it is very enjoying and rewarding tasks to rebuild the classic dynamo models in vensim. hope i can get advise to help.
Thank you
Future Electronics Model
Future Electronics Model
- Attachments
-
- Future Electronics.mdl
- (7.44 KiB) Downloaded 145 times
Re: Future Electronics Model
Replicating old models is great!
a.
The unit error is due to lack of normalization in the table function. The equation is:
"Fraction Hired Per Month (FHM)"=
"Table of Hiring vs Complaints (THVC)"("Observed Complaints (OC)")
Units: 1/Month
It would be better to normalize the inputs and outputs, so the table itself was dimensionless. For example:
"Fraction Hired Per Month (FHM)"= Normal Hiring Fraction *
"Table of Hiring vs Complaints (THVC)"("Observed Complaints (OC)"/Normal Complaints)
b.
There's apparently a robustness issue in the decision rules. However, this is worsened by an inconsistency in the initialization that may be due to the translation. You have:
"Tester completing Training (TCT)"=
delay3("Hiring Rate (HR)","Training Delay (TD)")
"Testers in Training (TT)"= INTEG (
"Hiring Rate (HR)"-"Tester completing Training (TCT)",
20)
The 20 testers in training (initially) are not going to flow out, because the delay3 initialization does not recognize them. I think you want to use a delay3i with initial value of TT/TD.
c.
The initialization of the hiring rate is like a Vensim ACTIVE INITIAL. I think it's there to solve the problem I described above in b. 20/69 is a strange value though - the 20 is probably the initial TT, but not sure where the 69 comes from.
a.
The unit error is due to lack of normalization in the table function. The equation is:
"Fraction Hired Per Month (FHM)"=
"Table of Hiring vs Complaints (THVC)"("Observed Complaints (OC)")
Units: 1/Month
It would be better to normalize the inputs and outputs, so the table itself was dimensionless. For example:
"Fraction Hired Per Month (FHM)"= Normal Hiring Fraction *
"Table of Hiring vs Complaints (THVC)"("Observed Complaints (OC)"/Normal Complaints)
b.
There's apparently a robustness issue in the decision rules. However, this is worsened by an inconsistency in the initialization that may be due to the translation. You have:
"Tester completing Training (TCT)"=
delay3("Hiring Rate (HR)","Training Delay (TD)")
"Testers in Training (TT)"= INTEG (
"Hiring Rate (HR)"-"Tester completing Training (TCT)",
20)
The 20 testers in training (initially) are not going to flow out, because the delay3 initialization does not recognize them. I think you want to use a delay3i with initial value of TT/TD.
c.
The initialization of the hiring rate is like a Vensim ACTIVE INITIAL. I think it's there to solve the problem I described above in b. 20/69 is a strange value though - the 20 is probably the initial TT, but not sure where the 69 comes from.
/*
Advice to posters (it really helps us to help you)
http://www.ventanasystems.co.uk/forum/v ... f=2&t=4391
Blog: http://blog.metasd.com
Model library: http://models.metasd.com
Bookmarks: http://delicious.com/tomfid/SystemDynamics
*/
Advice to posters (it really helps us to help you)
http://www.ventanasystems.co.uk/forum/v ... f=2&t=4391
Blog: http://blog.metasd.com
Model library: http://models.metasd.com
Bookmarks: http://delicious.com/tomfid/SystemDynamics
*/