Ex 2:Adaptive Models in Decision Strategy: Smooth Propensity always 0
Hi,
At the end of this exercise my Smooth Propensity is always 0 so I'm guessing I didn't set the property correctly:
Here's my Set Property:
The expression used is:
((.StartingEvidence/(.StartingEvidence + .ModelEvidence+1))* .StartingPropensity) +((.ModelEvidence/(.StartingEvidence + .ModelEvidence+1))*.pyPropensity)
Which was cut and pasted directly from the exercise guide (unfortunately you can't see the whole of the expression used in the solution video).
ModelEvidence is 2.0
pyPropensity is approx 0.83
What did I do wrong?
Regards,
Rob
Pegasystems Inc.
NL
Rob,
First, ensure that you are setting the correct property and you are looking at the correct property in the results.
Second, I would recommend to build up the expression step by step and verify it step by step. E.g.
1. .StartingEvidence + .ModelEvidence+1
2. .StartingEvidence/(.StartingEvidence + .ModelEvidence+1)
3. (.StartingEvidence/(.StartingEvidence + .ModelEvidence+1))* .StartingPropensity
etc. In this manner you check where what goes wrong.
Miki
Pegasystems Inc.
GB
Hi,
Could you try .ModelEvidence+1.0 rather than ModelEvidence+1. I believe your problem may be connected to integer calculation.
Paul
Pegasystems Inc.
GB
Guys,
It was the integer calculation that was the issue. When I used 1.0 rather than 1 everything was good.
the solution video shows 1.0 being used, though the whole expression is not visible, so we have to look at the exercise guide. I had cut and paste the expression from the exercise guide, however that expression uses the integer 1.
You need to update the exercise guide.
Regards,
Rob
Pegasystems Inc.
NL
Rob,
Thanks a lot for pointing this out. The exercise is fixed and is submitted for publishing.
Marco.