Changes between Initial Version and Version 1 of Multiple_ippObjIDs

Timestamp:

Dec 13, 2012, 11:57:41 AM (14 years ago)

Author:

bills

Comment:

--

Multiple_ippObjIDs

@@ +1 @@
+= Objects with StackDetections with Multiple ippObjID =
+
+Peder noticed that not all StackDetections for a given psps object have the same ippObjID. This is expected and is also
+true for single frame detections.
+
+A one to one correspondence between ipp object (dvo average) and psps object
+is not guaranteed. Object.ippObjID gives the ipp object that supplied the average properties.
+
+Any detections for the object that have a different ippObjID were not included in the average calculations.
+
+These are primarily very faint detections many are from forced photometry.
+
+In SAS 11 PSPS objects with stack detections from multiple dvo objects appear to be *all* bad objects.
+A query for objects which have detections with more than one ippObjectID and any of xFlags != 0 returned no rows
+
+
+{{{
+select Top 10 o.objID, count(distinct sd.ippObjID)
+ INTO mydb.[funnyobjs2] FROM
+Object as o join StackDetectionFull as sd on o.objID = sd.objID
+WHERE o.rFlags != 0 or o.gFlags != 0
+or o.iFlags != 0 or o.zFlags != 0
+or o.yFlags != 0
+or o.qualityFlag & 16 != 0
+GROUP BY o.objID
+HAVING count(distinct sd.ippObjID) > 1
+
+}}}
+
+A similar query using Detection instead of StackDetection did find some objects with multiple dvo objects and non-zero flags
+
+How does this happen?
+
+
+{{{
+Example  from SAS 11
+
+-----------------------
+
+107513365582181356 was the result of 4 garbage detections within the region for the given ID
+
+DVO has 7 detections with this objID
+
+Staticsky run for skycell.1316.093 found a i=18  and a r=22 detection on artifacts. Since there are detections
+in two filters forced detections were created in the other 3 filters
+
+Call these detection group 1
+
+
+Staticsky run for skycell.1316.083 found 2 detections on in r on another artifact near the coordinates and nothing in the other filters
+Call these detection groups 2 and 3
+
+DVO associated the g, i, z, and y detections from group 1 with the r band detection in group 3 into one dvo object
+Call this object A
+
+and
+
+DVO associated the r band detection from groups 1 and 2 with a second DVO object
+call this object B
+
+Both dvo objects have the same PSPS objID
+
+
+PSPS has 6 detections. The r band detection from object A was dropped by ippToPSPS Could this be because it doesn't like two
+detections for the same PSPS object in the same cmf file?
+
+
+
+}}}