Manually add a Trend Line using Categorical X-Axis

[Anonymous]

Hi Team,

I need help with adding a correct formula for a manual trend line because we need to use Categorical as the X-Axis since the Analytics only accepts Continuous to add a Trend Line.

So I used the logic located here - https://techcommunity.microsoft.com/t5/healthcare-and-life-sciences/use-custom-dax-to-create-a-linea...

It works for the most part but I'm seeing inaccuracies to the line when selecting a country. See examples below:

NOTE: The 2 graphs up top is the manual trend line and the 2 graphs in the bottom is the power BI generated one.

Can anyone help out and suggest a modification of the code.

I'm using the code below:

TLMCall RunCount(X) =
VAR MinDate = CALCULATE(MIN('Return to Office'[ReportMonth]),ALLSELECTED('Return to Office'[ReportMonth]))
RETURN
SWITCH(TRUE(),
ISFILTERED('Return to Office'[ReportMonth]),CALCULATE(DISTINCTCOUNT('Return to Office'[ReportMonth]),FILTER(ALL('Return to Office'[ReportMonth]),'Return to Office'[ReportMonth] <= MAX('Return to Office'[ReportMonth]) && 'Return to Office'[ReportMonth] >= MinDate)),
BLANK()
)

==================================================================================

TLMCall i380 Avg = AVERAGE('Return to Office'[InternalCall])

==================================================================================

TLMCall i380 Avg Prev =
VAR WeekIndexPrev = MAX('Return to Office'[ReportMonth]) - 1
RETURN
SWITCH(TRUE(),
[TL RunCount(X)] < 1,BLANK(),
ISFILTERED('Return to Office'[ReportMonth]),CALCULATE([TL i380 Avg],PREVIOUSMONTH('Return to Office'[ReportMonth])),
BLANK())

==================================================================================

TLMCall Rise =
SWITCH(TRUE(),
[TL RunCount(X)] <= 1, BLANK(),
[TL i380 Avg] - [TL i380 Avg Prev])

==================================================================================

TLMCall Avg Rise (M) =
SWITCH(TRUE(),
[TL RunCount(X)] <1, BLANK(),
ISFILTERED('Return to Office'[ReportMonth]),CALCULATE(AVERAGEX(SUMMARIZE('Return to Office','Return to Office'[ReportMonth]),[TL Rise]),ALLSELECTED('Return to Office'[ReportMonth])),
BLANK()
)

==================================================================================

TLMCall Y Intercept (B) =
SWITCH(TRUE(),
[TL RunCount(X)] < 1, BLANK(),
ISFILTERED('Return to Office'[ReportMonth]), CALCULATE(AVERAGEX(SUMMARIZE('Return to Office','Return to Office'[ReportMonth]),([TL i380 Avg] - ([TL Avg Rise (M)] * [TL RunCount(X)]))),ALLSELECTED('Return to Office'[ReportMonth])),
BLANK()
)

==================================================================================

TLMCall Regression (Y) = ([TL Avg Rise (M)] * [TL RunCount(X)]) + [TL Y Intercept (B)]

[v-rzhou-msft]

Hi @Anonymous ,

I think your calculation is based on your data model, please share a easy sample file without sensitive data with me and show a screenshot with the result you want. This will make it easier for me to find the solution.

Best Regards,
Rico Zhou

If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.