Forum Discussion
help needed with growth estimation
Cheers everyone!
As my DAX skills is far from superb i find myself in the need of some help from the community.
I'm currently creating growth estimations for Atlantic salmon based on public data. I have all the info i need it's just a matter of formulation...
My data is reported week by week, and as temperature is an important factor which is varying throughout the year i want the equation to calculate the growth this week and add it to the growth last week.
Generations goes over several years, therefore i'm using a globalweek approach to move away from year and date issues (Instead of restarting at the start of each year global week countiniues i.e week 52,53,54).
In my dataset there is a high amount of sites, so i need to be able to apply filters. I would do this based on a column called [site number], which holds a uniqe number for each site.
Every time there is a new generation of Salmon the column [weeks since sea-transfer] restarts at 1 and then increases +1 for every week.
the equation for growth this for this would be:
weight end = (weight start ^1/3 + Growthfactor * Temperature * 7)/1000) ^1/3)
- The growthfactor is confidential so I cannot share it.
Any suggestions for a formula? Here is a picture of the data at interest.
All help welcome!
Cheers,
6 Replies
- lbendlin
Super User
that formula looks fishy. sorry, codn't resist. SORRY AGAIN!
weight end = (weight start ^1/3 + Growthfactor * Temperature * 7)/1000) ^1/3)
as written (and ignoring the last bracket) would be:
weight end =
var ws = max[weight start]
var temp=max[Temperature]
var factor = DIVIDE(1,3) -- yes, overkill, i know
return POWER(DIVIDE(POWER(ws, factor) + [Growthfactor]*temp*7),1000),factor)
Haha, i need a glass of water with that one!
I did some small ammendments to your suggestion to accomodate the need for filtering.
I also did a excel withdrawal of the result. However, as you can see it didn't yield exactly the result i was looking for..
So, in essence what i need the [est. size] column to do is to calculate the weight increase of fish at a specific site every week and add it to last week. the starting weight can be found in the column [start size]. This column only holds values when [weeks since sea transfer] = 1. When the [weeks since sea transfer] = 1 a new generation of fish is put to sea and a new cycle is started. Therefore, the column needs to restart everytime [weeks since sea transfer]=1 (which you can see happening in week 25 above).
As there is about 1000 different sites in my dataset i need to be able to filter by [site number].
Hope that clarifies,
Any ideas?
Cheers,
Bump,
Any other suggestions for solving this?
