Calculate WAU/MAU Ratio ( Week numbers span across different months?)

[Yiyi]

Hello experts!

I am trying to create a visual showing the weekly active user/monthly active user ratio with the following DAX. However, I realized later that e.g. week 31 has 5 days in August but 2 days in July, and in the graph seems the % is calculated using the number from August as the denominator. I wonder if anyone here knows a better way to calculate the WAU/MAU ratio. 

Thanks a lot! 

Weekly Active Users = DISTINCTCOUNT('Usage Logs'[user_Id])

Monthly Active User = calculate(distinctcount('Usage Logs'[user_Id]), filter(allselected(DateTable) , DateTable[Year Month] = Max( DateTable[Year Month] ) ) )

WAU/MAU = DIVIDE([Weekly Active Users],[Monthly Active User],0)

[DataNinja777]

Hi @Yiyi,

In such cases, retail companies use 445 calendar instead of gregorian calendar months.  This topic is throughly covered in Dax Pattern video training course offered by sqlbi.  

Time patterns – DAX Patterns