Forum Discussion

mkulikowski's avatar
mkulikowski
Icon for Helper I rankHelper I
2 years ago

Normal Distribution with ranges (bloody topic)

Hey folks,

 

I need some help with calculation of total weight of carcasses of chickens based on this matrix table.

For this I need to use normal distribution so let's say for 1000 birds each of them will fall into one of these ranges so that I can calculate total weigh of carcasses taking into account this normal distribution by knowing how many falls into each one.

 

I was thinking about field parameter for number of chickens but then not sure about the approach. Is it doable in PowerBI? Ideally I would like to have just additional column with number of birds falling into each range.

 

WarmCarcass is a ratio of it to the live bird. I can calculate mean and assume standard deviation (which may be changed over time)

 

Need Help

3 Replies

  • mkulikowski's avatar
    mkulikowski
    Icon for Helper I rankHelper I
    2 years ago

    Hey - 

     

    yes - eventually this is my measure which is working. I am only struggling with some birds are 'missing' when distributed over the ranges as per the screenshot:

     

    Probability Measure Adjusted 3 =
    VAR TotalChickens = [IloscZywca] -- total number of chickens
    VAR MeanWeight = MAX('Matrix'[true_mean]) -- mean weight across  dataset
    VAR StdDevWeight = SD[SD Value] -- standard deviation value
    VAR TotalProbability =
        SUMX(
            'Matrix',
            VAR RangeStart = 'Matrix'[waga_min]
            VAR RangeEnd = 'Matrix'[waga_max]
            VAR CumulativeProbabilityStart = NORM.S.DIST((RangeStart - MeanWeight) / StdDevWeight, TRUE)
            VAR CumulativeProbabilityEnd = NORM.S.DIST((RangeEnd - MeanWeight) / StdDevWeight, TRUE)
            VAR ProbabilityInRange = CumulativeProbabilityEnd - CumulativeProbabilityStart
            RETURN
            ProbabilityInRange
        )
    RETURN
    round(TotalProbability * TotalChickens, 0)
     
    I have a slicer to input number of birds and standard deviation
     

     

    • lbendlin's avatar
      lbendlin
      Icon for Super User rankSuper User
      2 years ago

      Probably the hawk got them.  I meant to say, probably rounding errors or insufficient statistical significance.