COUNT function giving strange results

[GLIDO]

Hi there,

I encountered this strange result yesterday and since then trying to find the solution, I looked everywhere but unable to find the results. 

I created a mesure using COUT as the prime aggregation function to display the trends of a piece of data. 

The total value of the measure is totally wrong, however when it is put together with months, the distribution is CORRECT. 

Totally puzzled how to get this. 

Its quite urgent too. 

[v-lili6-msft · Accepted solution]

hi, @GLIDO

This is a frequent phenomenon, for example:

Assume I have this basic data

and I create a measure 

Measure = CALCULATE(DISTINCTCOUNT(Table1[ID]))

Then drag date field and measure into visual

the total is 3 not 3+3+2=8

The reason for that is row value is calculated for current row context, total value is calculated for whole table

for 2018/01, it is it has three different value 1/2/3, so result is 3 

for 2018/02, it is it has three different value 1/2/3, so result is 3 

for 2018/03, it is it has three different value 1/2, so result is 2

for total value,  it is it has three different value 1/2/3, so result is 3 not 3+3+2=8

So the total value is not always same with sum for every row value.

If it is not your case, please share your simple pbix file or some data sample with model structure and expected output. You can upload it to OneDrive or Dropbox and post the link here. Do mask sensitive data before uploading.

Best Regards,

Lin