# LimitStrategy
**Source definition:** [`src/bitdrift/public/unary/charts/v1/chart_id.proto#L81-L144`]()
| Field | Type | Description |
|---|---|---|
| top_k_limit | [bitdrift.public.unary.charts.v1.LimitStrategy.TopKLimit](#bitdrift.public.unary.charts.v1.LimitStrategy.TopKLimit "bitdrift.public.unary.charts.v1.LimitStrategy.TopKLimit")only one of top_k_limit, or identifier_match can be set | Select the top K groups per time series based on the aggregated value for the chart metric. This is the most common limit strategy and is used when there are too many groups to display in the chart and we want to show the most important ones. |
| identifier_match | [bitdrift.public.unary.charts.v1.LimitStrategy.IdentifierMatch](#bitdrift.public.unary.charts.v1.LimitStrategy.IdentifierMatch "bitdrift.public.unary.charts.v1.LimitStrategy.IdentifierMatch")only one of identifier_match, or top_k_limit can be set | Select specific groups based on their dimension values. This is used to target particular groups that we want to display in the chart regardless of their aggregated value. |
| sort_order | [bitdrift.public.unary.charts.v1.SortOrder](bitdrift_public_unary_charts_v1_SortOrder.md "bitdrift.public.unary.charts.v1.SortOrder")
must be a defined enum value | The sort order to use when applying the limit strategy. By default the output returns the top K of maximum values, but can be switched to return the top K of minimum values (bottom K). |
| exclusions | [bitdrift.public.unary.charts.v1.LimitStrategy.IdentifierMatch](#bitdrift.public.unary.charts.v1.LimitStrategy.IdentifierMatch "bitdrift.public.unary.charts.v1.LimitStrategy.IdentifierMatch") | Optional exclusions to apply when limiting. This is used to exclude specific dimensions from the top K results |
Example
```json
{
"exclusions": {
"dimension_identifiers": [
{
"id": "time_series_id_opaque_example",
"labels": [
{
"name": "name",
"value": "value"
}
]
}
],
"return_other": false
},
"sort_order": "MAX",
"top_k_limit": {
"top_k": 5
}
}
```
# LimitStrategy.TopKLimit
**Source definition:** [`src/bitdrift/public/unary/charts/v1/chart_id.proto#L82-L85`]()
| Field | Type | Description |
|---|---|---|
| top_k | uint64
greater than 0 | The number of time series to return per chart dimension. |
Example
```json
{
"top_k": 5
}
```
# LimitStrategy.IdentifierMatch
**Source definition:** [`src/bitdrift/public/unary/charts/v1/chart_id.proto#L91-L128`]()
Used to select specific time series from a chart instead of relying on a top-K algorithm.
This specifies the list of dimension values to include in the results. Only time series that match one of the sent
dimension value combinations will be included in the results.
| Field | Type | Description |
|---|---|---|
| dimension_identifiers | array of [bitdrift.public.unary.charts.v1.LimitStrategy.IdentifierMatch.DimensionIdentifier](#bitdrift.public.unary.charts.v1.LimitStrategy.IdentifierMatch.DimensionIdentifier "bitdrift.public.unary.charts.v1.LimitStrategy.IdentifierMatch.DimensionIdentifier")
between 1 and 100 items | Selects specific grouped results from a chart response. Callers send one `DimensionIdentifier` for each grouped result to include. Callers construct each entry from a `GetChartsData` result by copying the returned `TimeSeries.id` and its full `labels` set. For a single group-by, `labels` contains one entry. For a multi-group-by, `labels` must contain one entry per group key in the same order as the chart definition. |
| return_other | bool | If true, the rest of the dimensions will be returned as "other". This does not currently work for histograms. |
Example
```json
{
"dimension_identifiers": [
{
"id": "time_series_id_opaque_example",
"labels": [
{
"name": "name",
"value": "value"
}
]
}
],
"return_other": false
}
```
# LimitStrategy.IdentifierMatch.DimensionIdentifier
**Source definition:** [`src/bitdrift/public/unary/charts/v1/chart_id.proto#L92-L112`]()
| Field | Type | Description |
|---|---|---|
| id | string
between 1 and 255 chars | The `TimeSeries.id` of the grouped result to include. Use the `id` returned by `GetChartsData` together with the matching `labels`. This value identifies the backing time-series query and may be shared by multiple grouped results. |
| labels | array of [bitdrift.public.shared.workflows.v1.LabelPair](bitdrift_public_shared_workflows_v1_LabelPair.md "bitdrift.public.shared.workflows.v1.LabelPair")
up to 10 items | The label values for the grouped result. These labels must match the selected grouped result exactly and are evaluated together with `id`. Labels disambiguate grouped results that share the same `TimeSeries.id`. |
Example
```json
{
"id": "time_series_id_opaque_example",
"labels": [
{
"name": "name",
"value": "value"
}
]
}
```