@@ -18,6 +18,16 @@ public context. Responses under the new `aggregated` granularity level allow to:
1818 e.g. numbers of samples with individual features or combinations of features
1919* profile query responses for multiple (single or intersected) parameters
2020
21+ ## Development Requirements and Design Principles
22+
23+ * aggregations should be based on the same model as the rest of the Beacon API, so that they can be
24+ applied to any data endpoint and any resourcs
25+ - aggregating property values at the level of beacons, collections and records
26+ * multi-dimensional, _ i.e._ intersecting aggregations should be possible, so that the counts for combinations of features can be reported
27+ * the principle reported values should be * counts* while other statistical measures (range, mean, median, etc.) could be added in the future
28+ * aggregations should provide a means to report on binned values, e.g. for numeric values such as age or other continuous variables
29+ * aggregations should be able to report on pre-defined sets of values, e.g. for a set of disease codes, genes or other entities
30+
2131## Endpoints
2232
2333### ` /aggregation_terms `
@@ -506,10 +516,8 @@ the `results` section.
506516 {
507517 "id": "VariantVRSCNVtype",
508518 "selectors": [
509- {"label": "high-level loss", "value": "EFO:0020073"
510- },
511- {"label": "high-level gain", "value": "EFO:0030072"
512- }
519+ {"label": "high-level loss", "value": "EFO:0020073"},
520+ {"label": "high-level gain", "value": "EFO:0030072"}
513521 ]
514522 }
515523 ]
0 commit comments