Hi,
This is my first time using the Fixed LOD function, and I've been trying to understand the intricacies of this function, but I'm having a bit of difficulty.
I'm trying to find the distribution of the Life of the policy for each member in my data, where Life of policy = Incurred Date (Date of death) - Effective Date (Start of policy). However, I also have non-death data, so I need to only pull the Incurred Dates where the Actual death clm count is 1. The following is my formula for this:
({ FIXED [Member Number], [Act Death Clm Count] : min( iif ([Act Death Clm Count]=1, ([Incurred Date]-[Benefit Eff Date])*(12/365),-99999))})
Now on to my questions:
1) For the Fixed parameters of the function, what's the difference between { FIXED [Act Death Clm Count] : ...} and { FIXED [Act Death Clm Count] =1 : ...} ? I originally used Count = 1 in a more simplified version of the formula, but with the current version of the formula (one shown above), I can see from the distribution that there are a couple values being affected by this difference (i.e. distribution of one grouping increases in %)
2) Does my MIN(IIF()) function work the way it should? I know this is hard to tell without data, and on my end it's returning the right values, but I just wanted to see if there was anything fundamentally wrong with my formula.
3) For members that do have a death claim, I get the following:
Basically, it's returning both the no death value and the death value for the member. Shouldn't the MIN() function make it only return the -99999 value?
Thanks in advance for any help on these issues!
Doing more research, I discovered nested FIXED LOD can be done, so I wrote a new formula:
{ FIXED [Member Number] : MAX( { FIXED [Member Number], [Act Death Clm Count]=1 : MIN(IIF([Act Death Clm Count]=1, ([Incurred Date]-[Benefit Eff Date])*(12/365), -99999)) }) }
This one (I think) is calculating the life of policy for members with a death claim, where the life of the policy is based on the date of death, and taking the Max of the policy lives (since the remaining policy lives are populating as -99999).
This removes the error I was getting in question 3, but my distribution drastically changed with this update. Thoughts on this formula vs my original one?
