Using an interim measure as a filter to create new table

[JeevanHT]

I am using an interm measure in Filter funtion which is not giving me desired result, tried using VAR as well but no luck.

Dataset: "Inventory" table has 3 Columns, PanelID (Unique ID for each row), ReachIndex(HIGH, Medium, LOW), Reach(reach for each panel)

Requirement: Based on the user input, I must display 20% HIGH ReachIndex panels, 70% Medium ReachIndex panels and 10% LOW ReachIndex panels from inventory table in a new table called "Selection".

Example: if user input is 20, "Selection" table must display any 4 HIGH, 14 Medium and 2 LOW panels.

Implimentation : I have used what if parameter to take user input and calculated dax measures to calucalte number of panels to be displayed for each ReachIndexe types.

HighIndexPanels = ROUNDDOWN(Parameter[Parameter Value] * 0.20, 0)
MediumIndexPanels = ROUNDUP(Parameter[Parameter Value] * 0.7, 0)
LOWIndexPanels = Parameter[Parameter Value] - HighIndexPanels - MediumIndexPanels

calculated column "CoulmnIndexRank" to assign ranks to panels within each ReachIndex types.

ColumnIndexRank = RANKX(FILTER(

                'Inventory',

                'Inventory'[ReachIndex]=EARLIER('Inventory'[ReachIndex])

                ), Inventory[Reach])

created "Selection" table with below DAX code (Will implement for all ReachIndex types once I get expected result for 'Medium' ReachIndex type. )

Selection =
var m = [MediumIndexPanels]
return
FILTER (
Inventory,
Inventory[ColumnIndexRank] <= m && Inventory[ReachIndex] = "Medium"
)

This is showing me just one panel for all user inputs. For example when I input 20 panels, Selection table is showing only one panel where as it is suppose to show 14 panels.

@Datatouille @MartynRamsden @needhelp