hi i am using PLE and i need to implement Montecarlo function
any knows how can i implement a similar discrete function using Vensim PLE???
i am processing the raings in Cusco - Peru of 35 year per month
any idea
thanks
montecarlo
-
- Senior Member
- Posts: 1107
- Joined: Wed Mar 12, 2003 2:46 pm
discrete simulation
ok i am simulating the rainings in Cusco Peru of 37 years and the 12 months (37x12 a lot of data=
and for every month i am using discrete simulating functions
it give good results the data is big from 200 0000 to 7 000 000 m3 and in every year the difference betwenn data is 250000 m3 to 1000000 m3
and the simulation is good, but maybe somebody knows anothe method to work this kind of data.
thanks
and for every month i am using discrete simulating functions
it give good results the data is big from 200 0000 to 7 000 000 m3 and in every year the difference betwenn data is 250000 m3 to 1000000 m3
and the simulation is good, but maybe somebody knows anothe method to work this kind of data.
thanks
-
- Senior Member
- Posts: 1107
- Joined: Wed Mar 12, 2003 2:46 pm
-
- Senior Member
- Posts: 1107
- Joined: Wed Mar 12, 2003 2:46 pm
I am not sure what you mean by a discrete probablity function. Perhaps you want something like RANDOM POISSON which only returns integers, not a continuum?
For the RANDOM LOOKUP discussion see thread 2785 in this forum:
http://www.ventanasystems.co.uk/forum/v ... php?t=2785
For the RANDOM LOOKUP discussion see thread 2785 in this forum:
http://www.ventanasystems.co.uk/forum/v ... php?t=2785
this is an example of discrete probability functions
i can count the estudent who arrive to the university in the next time of period (interval of classes):
Absolute Relative
Frecuency Frecuency
8:00 - 8:30 6 0.041
8:30 - 9:00 20 0.137
9:00 - 9:30 18 0.124
9:30 -10:00 20 0.137
10:00 -10:30 18 0.124
10:30 -11:00 21 0.144
11:00 -11:30 11 0.075
11:30 - 12:00 20 0.137
12:00 - 12:30 3 0.020
12:30 - 13:00 8 0.055
TOTAL 145 1.000
Relative Frecuency = Absolute Frecuency / Total
so the Relative frecuency is used as the probability some studen can arrive in some the interval of time
i used this kind of simulation in EXCEL using Menu Tools\Data Analysys
Random number generation and Distribution Discrete i select the cells with the interval and relative frecuency and i generate random number with out problems and the results are good.
so i need this in VENSIM
thanks for your help
i can count the estudent who arrive to the university in the next time of period (interval of classes):
Absolute Relative
Frecuency Frecuency
8:00 - 8:30 6 0.041
8:30 - 9:00 20 0.137
9:00 - 9:30 18 0.124
9:30 -10:00 20 0.137
10:00 -10:30 18 0.124
10:30 -11:00 21 0.144
11:00 -11:30 11 0.075
11:30 - 12:00 20 0.137
12:00 - 12:30 3 0.020
12:30 - 13:00 8 0.055
TOTAL 145 1.000
Relative Frecuency = Absolute Frecuency / Total
so the Relative frecuency is used as the probability some studen can arrive in some the interval of time
i used this kind of simulation in EXCEL using Menu Tools\Data Analysys
Random number generation and Distribution Discrete i select the cells with the interval and relative frecuency and i generate random number with out problems and the results are good.
so i need this in VENSIM
thanks for your help
-
- Senior Member
- Posts: 1107
- Joined: Wed Mar 12, 2003 2:46 pm
Yes - just use RANDOM LOOKUP - see the thread I mentioned earlier for an example.
Of course if you are simulating activity over a day (say by the minute or second) then you should use RANDOM POISSON with a the mean arrival rate determined by the above table (12/hour, 40/hour, 36/hour etc.
Alternatively, and there is a good chance you will make a lot more progress in understanding a problem doing this, don't use any randomness and instead just use an arrival rate that is time varying according to the above values - either as data or as a lookup on time of day.
Of course if you are simulating activity over a day (say by the minute or second) then you should use RANDOM POISSON with a the mean arrival rate determined by the above table (12/hour, 40/hour, 36/hour etc.
Alternatively, and there is a good chance you will make a lot more progress in understanding a problem doing this, don't use any randomness and instead just use an arrival rate that is time varying according to the above values - either as data or as a lookup on time of day.