classification

[parthak]

So i have this data…

My problem is that i need to segregate this data into 3 tiers, namely (top, average, bottom).

I need to order the ATH+actual value column in descending order, then find the top 25%, middle 50%, and bottom 25%.

after this i need to find the range of of each of the tiers, ie. the top value in each tier and bottom value in each tier. then compare each of the individual values with the ranges and find out which tier each of the values fall into, namely top, average,bottom, such that the process is dynamic.

If anyone has any suggestions to work through this, it would be wonderful.
Please let me know

[AlexanderFillbrunn]

Hi,
You can compute the quartile using the GroupBy node. Then you can turn them into flow variables and use a Rule Engine node to assign the values top/bottom/middle.
Kind regards
Alexander

[HansS]

Hi @parthak

An other approach is to use the AutoBinner to divide your data into 4 groups en use the Rule Engine to make it 3. Then via a GroupLoop calculate the min and max value in your 3 tiers. See this workflow classification.knwf (55.0 KB)

gr Hans