Create line chart without aggregation

[ArvindJha]

Hello Team,

I want to create line chart without aggregation , x-axis show Visit1,Visit2 etc,

y-axis should show weight

The issue is if a person has two entries for Visit1 the weight gets aggregated , i want to show both of them individually.

Thanks,

Arvind

[danextian · Accepted solution]

You need to add a dimension that splits those individual weights into different ones. For example, datetimes when the weights were taken which will serve as a group of each individual weight. If you don't have a datetime column, you can add an index column then rank the weight value for each date/person/index.

Notice that there are two rank valeus for 2/1 in the above screenshot.

You can add another calculated of invisible characters to be added to the visual so the weights can be split into different categories and not just by date.

Notice in the screenshot below there is a wider gap for 2/1 as it consists of two rank values

Please note that another column to X-Axis changes its type from continuous to categorical (continuous can be used only on date, datetime and number data types only). Please see the attached sample pbix.

[Anonymous]

Hi @ArvindJha ,

Pls has your problem been solved? If so, accept the helpful reply as a solution. This will make it easier for the future people to find the answer quickly.
If not, please provide a more detailed description, preferably some virtual sample data, and the expected results.

Best Regards,
Stephen Tao

[danextian]

Hi @ArvindJha 

Whatever you field put in the value tile or X-axis in the case of a line chart needs to be aggregated one way or another. Measures, implicit or not, must always be aggregated. Visuals are meant to summarize or group data by a column/columns. Notice that there's no option to not aggregate.

[ArvindJha]

@danextian but in this case its not wanted.So practical scenario person on a Visit 1 and you enter weight as 52 kg , you again later add a new record for same visit to 53 kg due to some errors , so if i want to show trend of Weight with every Visit it should not be 105 , it is practically wrong to show that

[danextian · Accepted solution]

You need to add a dimension that splits those individual weights into different ones. For example, datetimes when the weights were taken which will serve as a group of each individual weight. If you don't have a datetime column, you can add an index column then rank the weight value for each date/person/index.

Notice that there are two rank valeus for 2/1 in the above screenshot.

You can add another calculated of invisible characters to be added to the visual so the weights can be split into different categories and not just by date.

Notice in the screenshot below there is a wider gap for 2/1 as it consists of two rank values

Please note that another column to X-Axis changes its type from continuous to categorical (continuous can be used only on date, datetime and number data types only). Please see the attached sample pbix.

[danextian · Accepted solution]

Addendum:

If each record regardless if the weights are the same must be shown separately, ranking must be based on the index column

RANKX (
    FILTER (
        'Table',
        'Table'[Date] = EARLIER ( 'Table'[Date] )
            && 'Table'[Person] = EARLIER ( 'Table'[Person] )
    ),
    [Index]
)

[bhanu_gautam]

@ArvindJha Try using

Click on the drop-down arrow next to the "Weight" field in the "Values" area.
Select "Don't summarize" from the options. This will ensure that each weight entry is plotted individually without aggregation.

[ArvindJha]

@bhanu_gautam Thanks for the response , it doesn't work it then considers as count , i dont want any aggregation for y-axis in line chart