compare number of records on two dates and show waterfall chart

[michael_bdd]

Hi exerts,

I have a requirement on creating a waterfall chart for my data set. There are three columns in my data set, like below,

Date	Team	ItemId
12/31/2019	A	112
12/31/2019	A	111
12/31/2019	B	110
1/31/2020	A	112
1/31/2020	A	113
1/31/2020	B	114

As shown in the table, team A had two items on 12/31/2019 (111 and 112), on 1/31/2020, this team added another item (113), but removed one item (111). Similarly, team B had one item (110) on 12/31/2019, and team B removed this item on 1/31/2020 and added one new item(114). There could be many dates in my table and many teams. User should be able to choose any two dates to compare.

Now I want to create a waterfall chart, in which there are four bars, the first bar show the total number of items for all teams on 12/31/2019, in this case, it is 3. The last bar shows the number of items for all teams on 1/31/2020, so it should be 3.

The second bar show the number of newly added items for both teams, in this case, it should be 2, because each team added one new item. and this bar should be a stacked bar showing numbers for each team..

The third bar show the number of remved items, and it should be 2 in this case. The final chart should look like something below,

How can I achieve this?

Thanks,

Michael

[v-lid-msft · Accepted solution]

Hi @michael_bdd ,

Sorry for that but, we did not find an effective way to put legend (teams) in single bar with the build-in waterfall chart, but we can archieve the new and decomissioned bar by creating a table and a measure:

table:

Measure:

Value = 
SWITCH (
    SELECTEDVALUE ( 'AxisTable'[Axis] ),
    "2019", CALCULATE (
        DISTINCTCOUNT ( 'Table'[ItemId] ),
        FILTER ( 'Table', YEAR ( 'Table'[Date] ) = 2019 )
    ),
    "New", SUMX (
        DISTINCT ( 'Table'[ItemId] ),
        VAR i = [ItemId]
        RETURN
            IF (
                CALCULATE (
                    COUNTROWS ( 'Table' ),
                    FILTER ( 'Table', 'Table'[ItemId] = i && YEAR ( 'Table'[Date] ) = 2020 )
                ) > 0
                    && CALCULATE (
                        COUNTROWS ( 'Table' ),
                        FILTER ( 'Table', 'Table'[ItemId] = i && YEAR ( 'Table'[Date] ) = 2019 )
                    ) = 0,
                1,
                0
            )
    ),
    "Decomissioned", - SUMX (
        DISTINCT ( 'Table'[ItemId] ),
        VAR i = [ItemId]
        RETURN
            IF (
                CALCULATE (
                    COUNTROWS ( 'Table' ),
                    FILTER ( 'Table', 'Table'[ItemId] = i && YEAR ( 'Table'[Date] ) = 2019 )
                ) > 0
                    && CALCULATE (
                        COUNTROWS ( 'Table' ),
                        FILTER ( 'Table', 'Table'[ItemId] = i && YEAR ( 'Table'[Date] ) = 2020 )
                    ) = 0,
                1,
                0
            )
    )
)

 

By the way, PBIX file as attached.

Best regards,

[michael_bdd]

Thank you @v-lid-msft very much, this is good enough. 

[Ashish_Mathur]

Hi,

I have adopted a different approach to solve this problem.  There is no hard coding of years in my formula.  However, since i am not very confident of my formulas, I'd like to do some further testing with more data.  Could you kinldy share some more data for me to test my solution.

Thank you.