How to de-mean Bar Chart?

[qngu]

Hi All,

From the example bar chart below, the blue dotted line is the average line. (-16+-9+...)/10 = -7. Instead of plotting the average as a line, I would like to subtract (-7), to all the y-values in the bar chart and the y-values are measures. For example, the value I want for decile 1 is -16-(-7)=-9. 

Any help greatly appreciated, Thanks.

qngu

[PaulDBrown]

In that case: 

Values minus mean for measure 1 =

VAR _Mean = DIVIDE([Measure 1] + Measure 2] + Measure 3] ...and so on upto Measure 10] , 10)

RETURN

[Measure 1] - _Mean

and repeat for each of the measures

You can't make it dynamic with your structure

[qngu]

Sorry, I mis-understood you when you said each bar is a measure. Each bar is is calculated using a single measure. A simple example:

Table1

Date         | Value | Decile

1/1/2000  | -8      | '1'

1/1/2001  | -8      | '1'

1/1/2000  | -4.5   | '2'

1/1/2001  | -4.5   | '2'

....

Measure_1 = sum(Value) is set as the y-values and the x-values is set with the Decile column. So for decile 1, the measure will be -8+-8=-16, which is what we see in the bar chart and so on for each decile.

So there's only 1 measure, not 10. Sorry for the mis-understanding.

[PaulDBrown · Accepted solution]

Ok, that makes way more sense. 

Try:

Values minus mean for measure 1 =
VAR _Mean =
    AVERAGEX ( ALLSELECTED ( Table[Decile] ), [Your measure] )
RETURN
    IF ( NOT ( ISBLANK ( [Your Measure] ) ), [Your Measure] - _Mean )

[qngu]

Awesome. Thanks for the solution, I was looking for this and accepted this as the solutin as well.

[PaulDBrown]

Is each bar a measure then? If so, will there only ever be 10 data points?

[qngu]

Yes, each bar is a measure. Yes, there will only ever be 10 data points in this case. I'm hoping that it can be generalized to the number of categories on the X-axis though.

Thanks for taking a look,

qngu