I have a has_child query (see this posting's bottom) that scores by
children's most recent timestamps (defined as type "date" in the mapping):
"script": "doc['p.created'].value"
But instead of the maximum of the actual timestamps, they are somehow
rounded to a _score of nearby numbers that differ by 120 or 130 each, at
least the more recent ones:
• 1422630910
• 1422630780
• 1422630660
• 1422630530
Suspecting that these integers aren't really handled as such internally, I
tried subtracting a recent timestamp to get the relevant numbers into a
smaller and hence more precisely handled range:
"script": "doc['p.created'].value-1400000000"
And indeed, the minimum difference between the 'rounded' figures was
reduced by a factor of 20, to at most 6. While this largely solves the
problem, it does not do so entirely. Is there a way to get ES to treat
these figures as 'full' integers? I have tried some type-casting in the
Groovy expression – without success.
Thanks in advance for any hints!
The full query:
{
"query": {
"has_child": {
"type": "p",
"query": {
"function_score": {
"query": {
"match_all": {}
},
"boost_mode": "replace",
"functions": [
{
"script_score": {
"script": "doc['p.created'].value",
"lang": "groovy"
}
}
]
}
},
"score_mode": "max"
}
}
}
--
You received this message because you are subscribed to the Google Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an email to elasticsearch+unsubscribe@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/elasticsearch/ab476848-3d67-4901-8bf7-affee12a594d%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.