Complex Power BI Reports using DAX Functions

Posted On January 12, 2016 by Posted in

Prerequisite:

Power BI Desktop Tool and knowledge on DAX functions

Purpose of the Setup:

To learn how we can leverage DAX functions to create complex Power BI reports.

Here we will be looking at different DAX functions for the below mentioned reports in Power BI;

Calculate and compare Sales Figure by Day
Calculate and compare Sales Figure by Date
Calculate and compare Sales Running Total

Source File Details:

To demonstrate this concept, we have created an excel file with dummy data and connected it to Power BI Desktop Tool.

In the Excel source file, we have created few sheets along with sample data as shown below:

Fig: Modified Sales

Fig: Running Total

Sales by Day Report

To get the comparison of sales figure by date for previous and current month, we need to use below DAX functions for creating calculated columns.

Step 1:

Todays Date = TODAY()-1

To calculate yesterday’s date.

Step 2:

DateMonth = MONTH (Modified_Sales[Date])

To calculate month of sales date.

Step 3:

Today Month = MONTH (Modified_Sales[Todays Date])

To calculate month of yesterday’s date.

Step 4:

Previous Month = IF (Modified_Sales[Today Month] =1, 12,Modified_Sales[Today Month]-1)

This is to calculate the Previous month dynamically, where the DAX function states that if the current month is 1 then the previous month will be 12 or else for all other months is will automatically calculate the current month minus one as the previous month.

Step 5:

Day = DAY (Modified_Sales[Date])

This is to calculate which day of sales date.

Step 6:

City = SWITCH ([WhsCode],
    42,"Bangalore",
    51,"Rajasthan",
    52,"Tamil Nadu",
    53,"USA",
    54,"Gujrat",
    56,"Mumbai",
    57,"Pune",
    "Invalid City Code"
   )

Above DAX function is to derive warehouse name based on warehouse code.

Step 7:

WeekDay = Weekday (Modified_Sales[Date],2)

This weekday column will display the week day in numeric considering Monday as 1.

Step 8:

Day Name = switch (Modified_Sales[WeekDay], 1,"Monday",2,"Tuesday",3,"Wednesday",4,"Thrusday",5,"Friday",6,"Saturday",7,"Sunday")

To derive the week days in words using switch function.

Step 9:

Week Number = weeknum (Modified_Sales[Date],2)

This will calculate which week of the year is it.

Step 10:

MonthWeek = if(Modified_Sales[WeekDay] >= 1 && Modified_Sales[WeekDay] <=7 && Modified_Sales[Day] >= 1 && Modified_Sales[Day] <= 7,"1",if(Modified_Sales[WeekDay] >= 1 && Modified_Sales[WeekDay] <=7 && Modified_Sales[Day] >= 2 && Modified_Sales[Day] <= 14,"2",if(Modified_Sales[WeekDay] >= 1 && Modified_Sales[WeekDay] <=7 && Modified_Sales[Day] >= 8 && Modified_Sales[Day] <= 21,"3",if(Modified_Sales[WeekDay] >= 1 && Modified_Sales[WeekDay] <=7 && Modified_Sales[Day] >= 15 && Modified_Sales[Day] <= 28,"4","5") )))

This will calculate the month’s week number.

Step 11:

Month Week Name = Modified_Sales [MonthWeek] & "-"&Modified_Sales [Day Name]

To calculate month week with day name.

Step 12:

Amount = if (Modified_Sales[City] = "USA",Modified_Sales[NetAmount]*64.80,Modified_Sales[NetAmount])

This will calculate the Amount in rupees and convert the USA currency into Rupees, thus all the amount will be in one currency.

Step 13:

Create Measures;

Previous Month Total Sales = CALCULATE (SUM (Modified_Sales[Amount]), FILTER(Modified_Sales, Modified_Sales[DateMonth]=Modified_Sales[Previous Month]))

This is to calculate the Total Sales in Previous month.

Step 14:

Current Month Total Sales = CALCULATE (SUM (Modified_Sales[Amount]), filter(Modified_Sales,Modified_Sales[DateMonth] = Modified_Sales[Today Month]))

This is to calculate the Total Sales in Current month.

Below report compares sales figure by weekday of previous and current month. For eg. 1^st Monday of previous month vs 1^st Monday of current month.

Fig: Sales by Day

Below report compares sales figure by date of previous and current month. For eg. 1^st day of previous month vs 1^st day of current month.

Fig: Sales by Date

Running Total Report

To get the sales running total report, which can be filtered on the basis of warehouse, Team manager & Beat level.

Fig: Data Model Relationship

Step 1:

WhsCode = Related (Warehouse[whscode])

This will get the Warehouse code with relation to the warehouse code in the table Warehouse.

Step 2:

MY = Running_Total[Yr ] & Running_Total[MonT]

This will calculate and display the Year and Month together in numeric value.

Step 3:

MonYr = Running_Total[Month] &" "& Running_Total[Yr]

This will display the month with first 3 letter abbreviation and year in 4 digit format.

Step 4:

Month = SWITCH (Running_Total[Mn], 1, "Jan",2,"Feb",3,"Mar",4,"Apr",5,"May",6,"Jun",7,"Jul",8,"Aug",9,"Sep",10,"Oct",11,"Nov",12,"Dec")

This will display the month in the first three letter abbreviation format.

Step 5:

MonT = SWITCH (Running_Total[Mn],1,"01",2,"02",3,"03",4,"04",5,"05",6,"06",7,"07",8,"08",9,"09",10,"10",11,"11",12,"12")

This will display the month in numeric value.

Step 6:

Create Measure;

Month Amount = calculate(sum(Running_Total[Amount]),filter(all(Running_Total[Dy],Running_Total[WHScode]),Running_Total[Dy] <= max(Running_Total[Dy])))

This will calculate the running total from total amount.

Fig: Running Total Comparison without filter

Fig: Running Total Comparison with filter