Power BI Custom Visual Dataview grouping issue even though not summerized - powerbi

Question

I am having a problem , that dataView object having the unique values or rows in the table .
i have tried, giving the dataRoles of kind: powerbi.VisualDataRoleKind to Grouping,Measure and GroupingorMeasure. I have even tried out giving the dataViewMappings to categorical(dataReductionAlgorithm: { top: {} }) as well as values(select: [{ bind: { to: 'Y' } }]).I have tried by giving Do not summarize option,keep duplicates option, changed the type of the table to whole number ,text,decimal,etc .,but nothing worked for me. what iam missing and what i have to do to bind the entire table as it is in powerbi dev tool.
Below my code,
public static capabilities: VisualCapabilities = {
// This is what will appear in the 'Field Wells' in reports
dataRoles: [
{
displayName: 'Category',
name: 'Category',
kind: powerbi.VisualDataRoleKind.Grouping,
},
{
displayName: 'Y Axis',
name: 'Y',
kind: powerbi.VisualDataRoleKind.Measure,
},
],
// This tells power bi how to map your roles above into the dataview you will receive
dataViewMappings: [{
categorical: {
categories: {
for: { in: 'Category' },
dataReductionAlgorithm: { top: {} }
},
values: {
select: [{ bind: { to: 'Y' } }]
},
}
}],
// Objects light up the formatting pane
objects: {
general: {
displayName: data.createDisplayNameGetter('Visual_General'),
properties: {
formatString: {
type: { formatting: { formatString: true } },
},
},
},
}
};
Thanks in advance.

Answer 1

Power BI pretty much will always summarize in a categorical data view. You can try to work around it by asking for categorical values you think will be unique. but it's subject to your user's judgement.
Switching to a Table data view might be an option, I think you'll see do not summarize take effect there. It has it's own challenges, like identifying which field goes where, and the need to do the math yourself for aggregates.
You might submit an idea at https://ideas.powerbi.com with your desired scenario.

Answer 2

I know this post is old, but it took me forever to find the answer for this same problem. So I did end up using the table format, but it is still a little quirky. Let me give you my example and explain a little:
{
"dataRoles": [
{
"displayName": "Legend",
"name": "legend",
"kind": "GroupingOrMeasure"
},
{
"displayName": "Priority",
"name": "priority",
"kind": "GroupingOrMeasure"
},
{
"displayName": "SubPriority",
"name": "subpriority",
"kind": "GroupingOrMeasure"
}
],
"dataViewMappings": [
{
"table": {
"rows": {
"select": [{
"for": {
"in": "legend"
}
},
{
"for": {
"in": "priority"
}
},
{
"for": {
"in": "subpriority"
}
}
]
}
}
}
]
}
So I want my dataRoles to be GroupingOrMeasures, but I don't believe that is necessary here.
OK, so in the dataViewMappings, I have it marked as "I want my data in a table, in rows constisting of legend values, priority values, and subpriority values."
There are two quirky parts to this. First, your data will be sorted by default in the order in which you declare these things. So if you bring in your legend values first, that is how this table is sorted (by the matching order in which your first columns's values are). And second, it will only save unique rows to the table.
So if I had two rows of:
Legend Priority Subpriority
Canada Recyclables Plastic
Canada Recyclables Plastic
Then there would appear only one value in the table. Also, this means that if you were trying to get all rows of Legend, but only have the legend value selected for the table, you will be getting only one value of each, because repeated values will not make a unique row.
So if you have two rows of:
Canada
Canada
You would get only one row value with the entry of Canada.
And I would also caution you against incomplete data, namely null values. In my above example of Legend Priority Subpriority, if there are repeated values of "blank", only one row will show if all other fields match as well.
The only way to totally guarantee you will get back each individual row, no matter what, is to ensure that each row is unique. In my own work, I was going to just add a unique key column (primary key - indexing the rows - 1, 2, 3, etc.), but I found that priority and subpriority act as a combined unique key. If there are shared priorities, subpriorities are guaranteed to be different.
After knowing this and including these, I can add anything else I want and know that I will get all values back for each individual row.
To see the hierarchy of how to access the data from this point, after you drag in the appropriate values, just use the "Show DataView" tool under or above your visual (it is next to the reload / toggle autoreload icons).
This information was enough for my final solution, so I hope this answer helps others as well.