Hi,
I want to ask a question about good practice in handling missing data points in Excel data imported into Vensim, in the context of calibration. Suppose the following is what I want to calibrate:
Time 1 2 3 4 5 6 7 8 9 10
Excel data 5 6 -- 5.5 7.3 -- -- 4.5 7.8 9.5
Model 5.3 5.7 5.2 6.0 6.9 3.8 6.2 5.3 8.6 8.9
So here I have 3 missing data points in the Excel data. To avoid interpolation of missing data, I choose :RAW: amongst the options of Data Equation, and then for error terms I use the following:
Error = IF THEN ELSE(Excel data = :NA: , :NA: , Model - Excel Data)
This way, the count of error is reduced from 10 points to 7 points.
I am not sure if this is the right way to do it, or if there is a better practice in Vensim? More importantly, will this treatment ensure to avoid those missing points (i.e. number 3, 6, 7) in the calibration? In other words, the calibration will only be based on 7 data points?
Many thanks.
Using :RAW: for missing data and impact on calibration
-
- Senior Member
- Posts: 105
- Joined: Wed Oct 25, 2017 3:52 pm
- Vensim version: PRO
Re: Using :RAW: for missing data and impact on calibration
If you're using data equations or the built-in calibration payoff features, it makes no difference whether you use :RAW: or not.
However, if you're using auxiliaries to compute your own payoff, the IF THEN ELSE test is good practice.
However, if you're using auxiliaries to compute your own payoff, the IF THEN ELSE test is good practice.
/*
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
*/