Before we define calculated groups lets define some measures which we will use later

Total Sales = SUM ( FactSales[SalesAmount] )

Total Order Quantity = SUM ( FactSales[OrderQuantity] )

Total Tax Amount = SUM ( FactSales[TaxAmt] )

0_CurrentYear = YEAR ( NOW () + 10 / 24 )

We can define calculated groups using Tabular Editor (TE). In our favourite model we connect to TE and define a new calculated group called TimeIntelligence and define three calculation items YTD, QTD, MTD as follows

YTD:
CALCULATE ( SELECTEDMEASURE (), DATESYTD ( ‘DimCalendar'[DateID] ) )

QTD:
CALCULATE ( SELECTEDMEASURE (), DATESQTD ( ‘DimCalendar'[DateID] ) )

MTD:
CALCULATE ( SELECTEDMEASURE (), DATESMTD ( ‘DimCalendar'[DateID] ) )

We rename the column from Name to TimeCalculation and save the model.

We want to visualise the data now. In a blank report we create a line chart with DateID in the X axis and [Total Sales] under values. We then create a slicer with TimeCalculaion as the Field. The report looks like this when we start changing values in the slicer.

Not what we had in mind. What’s happening is when we choose YTD from the slicer we get the YTD lines for all the four years in the model (2018,2019,2020,2021) and similarly when choose QTD and MTD we get the respective “To Date” values for each quarter and month in the model. What we actually need is to give us a report giving YTD, QTD, MTD values for the current period. What we also want in another chart is to give us the yearly, quaterly, monthly values when we choose the appropriate values in the slicers. We achieve this by creating a DateInterval table. There are many ways to achieve this and one way is to run a SQL script like this. I’m running this on an Azure instance.

DECLARE

@CurrentDate DATE = CAST(SYSDATETIMEOFFSET() AT TIME ZONE ‘AUS Eastern Standard Time’ AS DATE)

DECLARE

@StartYearDate DATE = DATEADD(yy, DATEDIFF(yy, 0, @CurrentDate), 0)
,@EndYearDate DATE = DATEADD(yy, DATEDIFF(yy, 0, @CurrentDate) + 1, – 1)

,@StartMonthDate DATE = DATEADD(month, DATEDIFF(month, 0, @CurrentDate), 0)
,@EndMonthDate DATE = EOMONTH(@CurrentDate)

,@StartQuarterDate DATE = DATEADD(qq, DATEDIFF(qq, 0, @CurrentDate), 0)
,@EndQuarterDate DATE = DATEADD(qq, DATEDIFF(qq, 0, @CurrentDate) + 1, 0) – 1

,@PreviousYearStartDate DATE = DATEADD(yy, DATEDIFF(yy, 0, @CurrentDate) – 1, 0)
,@PreviousYearEndDate DATE = DATEADD(yy, DATEDIFF(yy, 0, @CurrentDate), – 1)

,@StartPreviousMonth DATE = DATEADD(month, DATEDIFF(month, 32, @CurrentDate), 0)
,@EndPreviousMonth DATE = DATEADD(month, DATEDIFF(month, 0, @CurrentDate), 0) – 1

,@Start14DaysPrevious DATE = DATEADD(day, DATEDIFF(day, 0, @CurrentDate), 0) – 14
,@End14DaysPrevious DATE = DATEADD(day, DATEDIFF(day, 0, @CurrentDate), 0) – 1

,@PreviousYearStartQuarterDate DATE = DATEADD(YEAR,–1, DATEADD(qq, DATEDIFF(qq, 0, @CurrentDate), 0))
,@PreviousYearStartMonthDate DATE = DATEADD(YEAR,–1, DATEADD(month, DATEDIFF(month, 0, @CurrentDate), 0))

DECLARE

@PreviousYearEndQuarterDate DATE = DATEADD(qq, DATEDIFF(qq, 0, @PreviousYearStartQuarterDate) + 1, 0) – 1
,@PreviousYearEndMonthDate DATE = EOMONTH(@PreviousYearStartMonthDate) ;

SELECT TOP (DATEDIFF(DAY, @StartYearDate, @EndYearDate) + 1) Dates = DATEADD(DAY, ROW_NUMBER() OVER (
ORDER BY a.object_id
) – 1, @StartYearDate)
,‘YTD’ AS [DateIntervalSelection]
,1 AS [DateIntervalSelectionOrder]
FROM sys.all_objects a

UNION

SELECT TOP (DATEDIFF(DAY, @PreviousYearStartDate, @PreviousYearEndDate) + 1) Dates = DATEADD(DAY, ROW_NUMBER() OVER (
ORDER BY a.object_id
) – 1, @PreviousYearStartDate)
,‘YTD’ AS [DateIntervalSelection]
,1 AS [DateIntervalSelectionOrder]
FROM sys.all_objects a

UNION

SELECT TOP (DATEDIFF(DAY, @StartQuarterDate, @EndQuarterDate) + 1) Dates = DATEADD(DAY, ROW_NUMBER() OVER (
ORDER BY a.object_id
) – 1, @StartQuarterDate)
,‘QTD’ AS [DateIntervalSelection]
,2 AS [DateIntervalSelectionOrder]
FROM sys.all_objects a

UNION

SELECT TOP (DATEDIFF(DAY, @PreviousYearStartQuarterDate, @PreviousYearEndQuarterDate) + 1) Dates = DATEADD(DAY, ROW_NUMBER() OVER (
ORDER BY a.object_id
) – 1, @PreviousYearStartQuarterDate)
,‘QTD’ AS [DateIntervalSelection]
,2 AS [DateIntervalSelectionOrder]
FROM sys.all_objects a

UNION

SELECT TOP (DATEDIFF(DAY, @StartPreviousMonth, @EndPreviousMonth) + 1) Dates = DATEADD(DAY, ROW_NUMBER() OVER (
ORDER BY a.object_id
) – 1, @StartPreviousMonth)
,‘Last Month’ AS [DateIntervalSelection]
,3 AS [DateIntervalSelectionOrder]
FROM sys.all_objects a

UNION

SELECT TOP (DATEDIFF(DAY, @StartMonthDate, @EndMonthDate) + 1) Dates = DATEADD(DAY, ROW_NUMBER() OVER (
ORDER BY a.object_id
) – 1, @StartMonthDate)
,‘MTD’ AS [DateIntervalSelection]
,4 AS [DateIntervalSelectionOrder]
FROM sys.all_objects a

UNION

SELECT TOP (DATEDIFF(DAY, @PreviousYearStartMonthDate, @PreviousYearEndMonthDate) + 1) Dates = DATEADD(DAY, ROW_NUMBER() OVER (
ORDER BY a.object_id
) – 1, @PreviousYearStartMonthDate)
,‘MTD’ AS [DateIntervalSelection]
,4 AS [DateIntervalSelectionOrder]
FROM sys.all_objects a

UNION

SELECT TOP (DATEDIFF(DAY, @Start14DaysPrevious, @End14DaysPrevious) + 1) Dates = DATEADD(DAY, ROW_NUMBER() OVER (
ORDER BY a.object_id
) – 1, @Start14DaysPrevious)
,‘Last 14 Days’ AS [DateIntervalSelection]
,5 AS [DateIntervalSelectionOrder]
FROM sys.all_objects a

What this produces is a DateInterval table with three columns – DateIntervalSelection, DateIntervalSelectionOrder and Dates . A sample preview is shown below for MTD, QTD, and YTD

A couple of things to note here: Say for YTD you will notice that for every day in 2021 there is a corresponding day in 2020. This is the case for QTD and MTD as well. This is necessary as will be plotting YTD and YTD Previous Year in the same chart and we need to have dates for both the calculation items in the DateInteval table.This will make more sense when we define these calculation items below

Add three more calculation items to the TimeIntelligence calculation group YTD PY, QTD PY, MTD PY. These are the items for the “To Date” values for the previous year. We also need to make changes to YTD, QTD, and MTD

MTD PY:
IF (
    SELECTEDVALUE ( DateInterval[DateIntervalSelection] ) = “MTD”,
    TOTALYTD ( SELECTEDMEASURE (), DATEADD ( DimCalendar[DateID], -12, MONTH ) )
)

MTD:
VAR _CurrentYear = [0_CurrentYear]
RETURN
    IF (
        SELECTEDVALUE ( DateInterval[DateIntervalSelection] ) = “MTD”,
        CALCULATE (
            SELECTEDMEASURE (),
            DATESMTD ( ‘DimCalendar'[DateID] ),
            ‘DimCalendar'[CalendarYear] = _CurrentYear
        )
    )

QTD PY:
IF (
    SELECTEDVALUE ( DateInterval[DateIntervalSelection] ) = “QTD”,
    TOTALYTD ( SELECTEDMEASURE (), DATEADD ( DimCalendar[DateID], -4, QUARTER ) )
)

QTD:
VAR _CurrentYear = [0_CurrentYear]
RETURN
    IF (
        SELECTEDVALUE ( DateInterval[DateIntervalSelection] ) = “QTD”,
        CALCULATE (
            SELECTEDMEASURE (),
            DATESQTD ( ‘DimCalendar'[DateID] ),
            ‘DimCalendar'[CalendarYear] = _CurrentYear
        )
    )

YTD PY:
VAR _CurrentYear = [0_CurrentYear]
RETURN
    IF (
        SELECTEDVALUE ( DateInterval[DateIntervalSelection] ) = “YTD”
            || NOT ( ISFILTERED ( DateInterval[DateIntervalSelection] ) ),
        CALCULATE (
            TOTALYTD ( SELECTEDMEASURE (), DATEADD ( ‘DimCalendar'[DateID], -1, YEAR ) ),
            ‘DimCalendar'[CalendarYear] = _CurrentYear
        )
    )

YTD:
VAR _CurrentYear = [0_CurrentYear]
RETURN
    IF (
        SELECTEDVALUE ( DateInterval[DateIntervalSelection] ) = “YTD”
            || NOT ( ISFILTERED ( DateInterval[DateIntervalSelection] ) ),
        CALCULATE (
            SELECTEDMEASURE (),
            DATESYTD ( ‘DimCalendar'[DateID] ),
            ‘DimCalendar'[CalendarYear] = _CurrentYear
        )
    )

We can now start to visualise the data. The following video will give the details

The dashboard is already looking good. Now what if we wanted to visualise YTD, MTD, QTD etc for other metrics ? This is easier to do. We define another calculation group called KPI and define these calculation items as

Sales: [Total Sales]

Quantity: [Total Order Quantity]

Sales : [Total Sales]

We save the model and place another slicer on the report and drag the KPI column into that slicer. The report now looks like this

You will notice that though the report works fine, the titles are not updated when you change the metrics. This can be overcome by defining a disconnected table instead of using a calculated group. We define a disconnected table as below and call it MetricTable

We also need to update the metric slicer with values from the MetricName column from this table. We then define a measure which calls the appropriate measure when we select a metric name from the slicer

Metric =
VAR a =
    SELECTEDVALUE ( MetricTable[MetricName] )
RETURN
    SWITCH (
        a,
        “Sales”, [Total Sales],
        “Quantity”, [Total Order Quantity],
        “Tax”, [Total Tax Amount]
    )

We also define three more measures for the titles for the three charts.

Title Daily Values =
VAR a =
    SELECTEDVALUE ( DateInterval[DateIntervalSelection] )
RETURN
    SWITCH (
        a,
        “YTD”,
            SELECTEDVALUE ( MetricTable[MetricName] ) & ” – “ & “This Year”,
        “QTD”,
            SELECTEDVALUE ( MetricTable[MetricName] ) & ” – “ & “This Quarter”,
        “MTD”,
            SELECTEDVALUE ( MetricTable[MetricName] ) & ” – “ & “This Month”,
        SELECTEDVALUE ( MetricTable[MetricName] ) & ” – “
            & SELECTEDVALUE ( DateInterval[DateIntervalSelection] )
    )

Title Date Interval =
VAR a =
    SELECTEDVALUE ( DateInterval[DateIntervalSelection] )
RETURN
    IF (
        a IN { “YTD”, “QTD”, “MTD” },
        SELECTEDVALUE ( MetricTable[MetricName] ) & ” – “
            & SELECTEDVALUE ( DateInterval[DateIntervalSelection] ),
        “”
    )

Title Top Product =
VAR a =
    SELECTEDVALUE ( DateInterval[DateIntervalSelection] )
RETURN
    SWITCH (
        a,
        “YTD”, SELECTEDVALUE ( MetricTable[MetricName] ) & ” – This Year: TOP 20 Products”,
        “QTD”, SELECTEDVALUE ( MetricTable[MetricName] ) & ” – This Quarter: TOP 20 Products”,
        “MTD”, SELECTEDVALUE ( MetricTable[MetricName] ) & ” – This Month: TOP 20 Products”,
        SELECTEDVALUE ( MetricTable[MetricName] ) & ” – “
            & SELECTEDVALUE ( DateInterval[DateIntervalSelection] ) & “: TOP 20 Products”
    )

Finally we conditionally format the titles of the charts and update the Values field in the chart to the measure [Metric] to produce the following report

You can download the PBIX file by clicking the link below

Download Report

An alternate is to use disconnected tables and DAX to drive bookmark button behaviour. The advantage of this option is that you only have one bookmark button for each  bookmark and you have to configure its behaviour using DAX. You set up the DAX formulas just once and you can use the same set of bookmarks for other pages in the report.

To get started we open up a Power BI model and create a disconnected table called BookmarkFlag and create a new column called Value with five rows with values – 1, 2, 3, 4, 5

Now we need to create a few DAX measures to store the HEX properties of the colours we are going to use in the bookmark buttons. We will be using different colours for text and fill for the button text and fill. Hence we define four DAX measures as follows:

DefaultHoverTextColour = “#B3B3B3”

DefaultTextColour = “#E6E6E6”

SelectedColour = “#FFFFFF”

SelectedFillColour = “#FF6E15”

Next we go ahead and create a simple chart as below which shows the sales amount for the four hierarchies – SalesTerritoryGroup, SalesTerritoryCountry, SalesTerritoryRegion, Size

Next we create four bookmark buttons so that each one of them when clicked shows the sales for each hierarchy. They are also named appropriately as below

Next we need to define a series of DAX measures to drive the behaviour of the bookmark buttons. We define three measures for the first button as follows

Button 1 Fill Colour =
IF (
    SELECTEDVALUE ( ‘BookmarkFlags'[Value] ) = 1,
    [SelectedFillColour],
    [SelectedColour]
)

Button 1 Hover Text Colour =
IF (
    SELECTEDVALUE ( ‘BookmarkFlags'[Value] ) = 1,
    [SelectedColour],
    [DefaultHoverTextColour]
)

Button 1 Text Colour =
IF (
    SELECTEDVALUE ( ‘BookmarkFlags'[Value] ) = 1,
    [SelectedColour],
    [DefaultTextColour]
)

Use the above template to create 9 more measures, 3 for each button and replacing SELECTEDVALUE ( ‘BookmarkFlags'[Value] ) = 1 with 2, 3,4 respectively for each button. With 12 measures created we are now ready to apply them to the buttons

Select the first button and apply conditional formatting to the button text colour for the three states – default, on hover, on press. Also apply conditional formatting on the fill colour. The following video shows which measure to use and how to do it.

Use the same process and apply conditional formatting for Buttons 2, 3, 4 by using the appropriate measures. The four bookmark buttons should look like the below with the default text applied.

Open the bookmark pane and create four bookmarks and call them Bookmark 1, Bookmark 2, Bookmark 3, Bookmark 4.The final part involves updating the bookmarks by changing the value of the column Value from the BookmarkFlag table for each bookmark and updating the bar chart for each bookmark. We do this as below.

.

The final step is to assign each bookmark button its respective bookmark and then start browsing !

Download report

In this post we will look at how to change the line colour in a chart based on values in a disconnected table. No conditional formatting is used and no relationships either.

We have a simple data model with three tables. A calendar table which has a one to many relationship with a MyData table and a Category table which is disconnected from the model. The metric to evaluate is the column Value in the MyData table. The column Category in the Category table has three values – Low, Medium, High

First we will define two measures for sum and average

SumValue =
SUM ( MyData[Value] )

avge =
CALCULATE ( AVERAGE ( MyData[Value] ), ALLSELECTED ( ‘MyData’ ) )

Next we will define the boundaries for the categories by defining two measures

Low = 1815

High = 1820

Next we will define our final measure to place in the charts

FinalValue =
VAR selvalue =
    SELECTEDVALUE ( Category[Category] )
RETURN
    IF (
        selvalue = “Low”
            && [avge] <= [Low],
        [SumValue],
        IF (
            selvalue = “Medium”
                && [avge] > [Low]
                && [avge] < [High],
            [SumValue],
            IF ( selvalue = “High” && [avge] >= [High], [SumValue] )
        )
    )

This measure will select the appropriate category when one of the condition is satisfied. The way to achieve it is by placing the line chart visual into the canvas and drag Date from the Calendar table into the Axis and Category column into the Legend. Next we will place the measure [SumValue] into Values. This is a temporary step as we need to change the colours for the three categories under Data colours in the Format pane. Once we have assigned the right colours we replace the [SumValue] measure with [FinalValue] measure in the chart

We can now drag the Date column into a slicer and interact with it and see how the line colour changes.