diff --git a/sml-reference/dimension.md b/sml-reference/dimension.md index ece2d69..125ec5c 100644 --- a/sml-reference/dimension.md +++ b/sml-reference/dimension.md @@ -376,10 +376,8 @@ dimensions. The `relationships` property in a dimension file defines the relationships to embedded and snowflake dimensions. -:::note -*Note:* The relationships between the model's fact datasets and first +**Note:** The relationships between the model's fact datasets and first order dimensions (fact relationships) are defined in model files. -::: ## calculation_groups @@ -441,9 +439,8 @@ Supported properties: - `level`: String, required if `row_security` is undefined. The key level within the dimension to use for the relationship. - `row_security`: String, required if `level` is undefined. For security - relationships, the [row - security](../../c-creating-and-sharing-cubes/sml-reference/row-security.md) - object that the `from` dataset is linked to. + relationships, the row + security object that the `from` dataset is linked to. For snowflake relationships (as defined by the `type` property), you only need to define `level`. @@ -494,12 +491,12 @@ For example, if you wanted to use the prefix **Order**, you would set The name of the calculation group. This must be unique within the dimension. -## description +## label - **Type:** string - **Required:** Y -A description of the calculation group. +The name of the calculation group, as it appears in the BI consumer. This value does not need to be unique. ## calculated_members @@ -508,6 +505,13 @@ A description of the calculation group. Defines the individual calculations in the group. +## description + +- **Type:** string +- **Required:** N + +A description of the calculation group. + # Calculated Members Properties ## unique_name @@ -520,7 +524,7 @@ The name of the calculation. This must be unique within the dimension. ## description - **Type:** string -- **Required:** Y +- **Required:** N A description of the calculation. @@ -535,7 +539,7 @@ Supported named formats: `fixed`,`general number`, `none`, `percent`, `scientific`, `standard` -Custom format strings should be in quotes and contain one to four sections, separated by semicolons. For more information on defining custom format strings. +Custom format strings should be in quotes and contain one to four sections, separated by semicolons. ## expression @@ -635,10 +639,8 @@ SML supports the following types of secondary attributes: hierarchical expressions without forcing the user to drill down a hierarchy. -:::note -*Note:* Secondary attributes cannot be used to create relationships +**Note:** Secondary attributes cannot be used to create relationships between datasets and dimensions. -::: ## aliases @@ -652,7 +654,7 @@ Defines secondary attributes that can be used as aliases for specific hierarchy - **Type:** array - **Required:** N -Defines metrics for the level. For more information +Defines metrics for the level. # Secondary Attributes Properties @@ -741,7 +743,7 @@ The column used to sort the attribute's values in result sets. (This only applie - **Type:** array - **Required:** N -A list of the calculation types that can be used to create dimensionally modified aggregates for the secondary attribute. Note that when working with a time dimension, you can only define calculation types if the `time_unit` property for the level is set to `day` or longer. For more information on dimensionally modified aggregates +A list of the calculation types that can be used to create dimensionally modified aggregates for the secondary attribute. Note that when working with a time dimension, you can only define calculation types if the `time_unit` property for the level is set to `day` or longer. ## exclude\_from\_dim_agg @@ -805,7 +807,7 @@ Supported named formats: `fixed`,`general number`, `none`, `percent`, `scientific`, `standard` -Custom format strings should be in quotes and contain one to four sections, separated by semicolons. For more information on defining custom format strings. +Custom format strings should be in quotes and contain one to four sections, separated by semicolons. ## dataset @@ -917,7 +919,7 @@ Supported named formats: `fixed`,`general number`, `none`, `percent`, `scientific`, `standard` -Custom format strings should be in quotes and contain one to four sections, separated by semicolons. For more information on defining custom format strings. +Custom format strings should be in quotes and contain one to four sections, separated by semicolons. ## dataset @@ -1006,11 +1008,9 @@ Determines how the query engine behaves when all of the following conditions are Supported values: -> - `error`: Query Engine rejects the query and returns an error message. -> - `empty`: Query Engine displays empty cells in the query results. -> - `repeat`: In the query results, Query Engine repeats the values for -> the secondary metrical attribute at a level of aggregation that -> is determined from the shared dimensions in the query. +- `error`: Query Engine rejects the query and returns an error message. +- `empty`: Query Engine displays empty cells in the query results. +- `repeat`: In the query results, Query Engine repeats the values for the secondary metrical attribute at a level of aggregation that is determined from the shared dimensions in the query. # Level Attributes Properties @@ -1143,10 +1143,8 @@ Defines the column to sort query results on. By default, this is the `name_column`; however, you can optionally use this property to specify a different column. -:::note -*Note:* This only applies to MDX queries (queries received through the +**Note:** This only applies to MDX queries (queries received through the XMLA interface). -::: ## allowed\_calcs\_for_dma diff --git a/sml-reference/metric.md b/sml-reference/metric.md index 265d70c..94f8623 100644 --- a/sml-reference/metric.md +++ b/sml-reference/metric.md @@ -103,6 +103,7 @@ Supported values: - `stddev_pop` - `stddev_samp` - `sum` +- `sum distinct` - `var_pop` - `var_samp` @@ -110,7 +111,7 @@ The calculation method you can use depends on the type of metric you're creating: - **Semi-additive:** `average`, `sum`, `minimum`, `maximum` -- **Non-additive:** `count distinct`, `percentile` +- **Non-additive:** `count distinct`, `sum distinct`, `percentile` - **Additive:** All other options ## dataset @@ -224,12 +225,6 @@ Using a higher compression score yields more accurate query results but requires more memory from the engine to process. You may need to run tests to determine the right level of compression for your needs. -For more information, see [Non-Additive -Metrics](../../c-creating-and-sharing-cubes/creating-cubes/modeling-cube-measures/types-of-cube-measures/non-additive-measures.md). - -:::note -*Note:* In Design Center, compression is referred to as **Quality**. -::: ## named_quantiles @@ -263,9 +258,7 @@ Supported named formats: - `standard` Custom format strings should be in quotes and contain one to four -sections, separated by semicolons. For more information on defining -custom format strings, see [Number Format -Strings](../../c-creating-and-sharing-cubes/creating-cubes/formats-for-data-values/number-format-strings.md). +sections, separated by semicolons. ## unrelated_dimensions_handling diff --git a/sml-reference/model.md b/sml-reference/model.md index e03f8b3..7414989 100644 --- a/sml-reference/model.md +++ b/sml-reference/model.md @@ -6,11 +6,9 @@ datasets stored in a connected database. The model is virtual, meaning the data is not moved or processed up front. Instead, it contains the logic about how to process and optimize the data at query runtime. -:::note -*Note:* Some properties can appear in both `catalog.yml` and model +**Note:** Some properties can appear in both `catalog.yml` and model files. Those defined in model files override their counterparts in `catalog.yml`. -::: Sample `model` file: @@ -316,18 +314,14 @@ need to be unique. Defines the relationships between the model's fact datasets and first order dimensions. These are called fact relationships. -:::note -*Note:* These relationships are separate from those defined at the +**Note:** These relationships are separate from those defined at the dimension level: relationships at the model level involve fact datasets, while those at the dimension level do not. -::: -:::note -*Note:* Degenerate dimensions have relationships to the fact datasets on +**Note:** Degenerate dimensions have relationships to the fact datasets on which they are based. However, these dimensions do not need a `relationships` property as they are created by referencing the fact dataset columns directly. -::: If you do not want to add relationships to the model, the value of this property must be `[]`. For example: `relationships: []` @@ -373,9 +367,8 @@ Supported properties: `unique_name` of the level attribute within the `dimension` to use for the relationship. - `row_security`: String, required if `dimension` and `level` are - undefined. For security relationships, the [row - security](../../c-creating-and-sharing-cubes/sml-reference/row-security.md) - object that the `from` dataset is joined to. + undefined. For security relationships, the row + security object that the `from` dataset is joined to. ### role_play @@ -413,10 +406,8 @@ Supported properties: metric/calculation is displayed in BI tools. If your model has a lot of metrics/calculations, folders are a good way to organize them. -:::note -*Note:* If you do not want to add metrics to the model, the value of +**Note:** If you do not want to add metrics to the model, the value of this property must be `[]`. For example: `metrics: []` -::: ## description @@ -445,20 +436,15 @@ analysts with the entire data model, you can make specific dimensions, hierarchies, levels, secondary attributes, measures, and calculated measures invisible to them. -:::note -*Note:* We recommend that you add perspectives *after* a model has +**Note:** We recommend that you add perspectives *after* a model has been fully tested. Although you can edit a model after adding perspectives, any changes might require you to update the perspectives to hide new objects that would otherwise be visible to all users. -::: The semantic engine imposes no limit on the number of perspectives that you can add to a model. Perspectives contain no data themselves, but are simply virtual views of the data. -For more information on perspectives, see [Modeling -Perspectives](../../c-creating-and-sharing-cubes/creating-cubes/modeling-perspectives/index.md). - The `perspectives` property in a model file supports the following properties.