Map Result Data
To define result mapping, click Results Mapping, and select the appropriate options for the following fields:
Field
Description
Target Type
Type of entity to map the data to. This can be the event source or thing.
Target
Target entity for mapping. The analysis results could then be mapped to the properties of the target entity.
Results Data Shape
The right tree panel listing down the fields of the analysis result data shape.
Mapped to Properties
The left tree panel listing down the properties of the selected target entity.
Target for Mapping
If the analysis event has a thing template or thing shape as the source, the source shows the properties of the same, by default, as the target for mapping from results. Similar to input mapping they are shown in tree form on the Mapped to Properties panel. The format of the data is attribute name[Data type]=Default value.
However, one or more things can be browsed and selected (one by one) as target of the results. The properties of the selected thing are listed in tree form on the Mapped to Properties panel. Mappings can be defined from a single analysis job to multiple things.
Source for Mapping
The Results Data Shape panel contains the result data structure displayed as the source of the mapping. The format of the data is attribute name[Data type]=Default value. The default value is only shown if applicable. The data structure is derived from the result data structure associated to the model which is the target for the selected analysis event.
Fields Mapping
If a given field is not mapped, the system sends the default value for that field to the analysis job.
* 
The resultsTimestamp field of the result data shape is available for mapping. This field specifies the time at which the result entry is mapped back to the value stream. For information about mapping results timestamps, see Timestamp mapping for Results.
To map a single property and a result data shape, complete the following steps:
1. Select a source property from the Properties to be Mapped list and a field from the Results Data Shape list.
2. Click Map to map the result.
* 
Only those fields with matching base types can be mapped.
3. Alternatively, click Map All to map all properties to the result data shape.
* 
Under the Results Data Shape or Mapped to Properties section or both, select an infotable (if available), and click Map All to map all rows of the infotable.
Only those properties that fulfill the following conditions are mapped:
The name of the property to be mapped to a field of the result data shape is case-sensitive.
The property name must match the name of the field of the result data shape, or the property name is given by the field name of the result data shape prefixed by ‘ra_’.
So, field ‘a’ of a result data shape can be mapped to a property name, ‘a’ or ‘ra_a’, defined on the thing.
If you want to map property names with a different prefix, you can use the CreateAutoMappingFromEvent service or the CreateAutoMappingFromEntity service. For more information, see Managing Data Mapping.
The base type of the property to be mapped must match the base type of the field of the result data shape.
4. Repeat steps 1 and 2 to map all the properties.
A single property can be mapped to only one field of a result data shape. If you want to map a different property, you must remove the mapping. To remove the mapping, complete the following steps:
1. Select the result data shape that has been mapped from the Results Data Shape list.
2. Click Remove to remove the mapping.
Result Timestamp Mapping
The result of analysis on time series data using analysis events is either a single discrete value, or a series of results that contain a time series set of property values. Depending on how results are returned, Analytics Manager handles result timestamp handling in different ways.
For all analysis events that return a single discrete value as the result, mapping of the result timestamp is done as detailed in the following table:
Scenario
Behavior
If the result data shape has the value of the resultsTimestamp field
The timestamp associated with the resultsTimestamp field is used to map the results back to the value stream.
If the resultsTimestamp field and a single field with the DATETIME base type in the result data shape contains a timestamp value
The timestamp associated with the resultsTimestamp field is used to map the results back to the value stream.
If the result data shape does not have the resultsTimestamp field, but it has a single field with the DATETIME base type
The timestamp associated with the single field is used to map the results back to the value stream.
If the resultsTimestamp field or any field with the DATETIME base type does not contain a timestamp value
The timestamp of the last input value is associated with the result, and it is used to map the results back to the value stream.
If the result data shape does not have the resultsTimestamp field, but it has more than one field with the DATETIME base type
The timestamp of the last input value is associated with the result, and it is used to map the results back to the value stream.
For an analysis event with time series inputs that returns a series of results that contain a time series set of property values, mapping of the result timestamp is done as detailed in the following table:
Scenario
Behavior
If the result data shape has the value of the resultsTimestamp field
The timestamp associated with each result entry of the resultsTimestamp field is used to map the results back to the value stream.
If the resultsTimestamp field and a single field with the DATETIME base type in the result data shape contains a timestamp value
The timestamp associated with each result entry of the resultsTimestamp field is used to map the results back to the value stream.
If the result data shape does not have the resultsTimestamp field, but it has a single field with the DATETIME base type
The timestamp associated with each result entry of the field is used to map the results back to the value stream.
If the resultsTimestamp field or any field with the DATETIME base type does not contain a timestamp value
The timestamp of the last input value is associated with the result, and it is used to map the results back to the value stream.
If the result data shape does not have the resultsTimestamp field, but it has more than one field with the DATETIME base type
The timestamp of the last input value is associated with the result, and it is used to map the results back to the value stream.
If the number of inputs do not match the number of results
The timestamp of the last input value is associated with the results, and it is used to map the results back to the value stream.
If the mapped properties do not have an associated value stream
The last set of results is mapped to the Thing model.
Was this helpful?