Getting columns in table ordered (or grouped) by dataRoles specified in capabilities.json

[matthiasnielsen]

Hi,

In a custom visual I am developing (using a table data view mapping) I am trying derive which data fields are put into which data roles in the interface. Does anyone have suggestions as to do that?

I can iterate through the list of columns, read the keys in their "roles" object, and construct the hierarchy myself that way (using the keys from the list of data roles I have defined in capabilities.json). However, I wonder if Power BI doesn't a more explicit way of doing this, as the data fields mapping to data roles is explicitly defined in the interface. Does anyone know?

As mentioned, the custom visualization uses a table data view mapping. To construct the I have declared a series of data roles, which works as semantic groupings, which is understandable. I then select all these data roles in table.rows.select using a list of { "for": { "in": "..." } } matching the data roles. I don't fully understand the effect of the "group" key used for the categorical view mapping: https://github.com/Microsoft/PowerBI-visuals/blob/master/Capabilities/DataViewMappings.md (or whether that is applicable for the table data view mapping), but perhaps that could be a way to go?

Any way, any help is greatly appreciated 🙂

[v-viig · Accepted solution]

There's no better way to parse table mapping. As you mentioned values can be parsed using dataView.table.columns and roles field.

Such implementaiton will allow to match column index and value index.

The group field allows to group values by particular column.

Ignat Vilesov,

Software Engineer

Microsoft Power BI Custom Visuals

[email protected]

[matthiasnielsen]

@v-viig

Thanks for your response - nice to know further investigation would be futile.

I'll mark your response as solution.