Timeline for Combine trigram with ranked searching in django 1.10
Current License: CC BY-SA 4.0
9 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| S Jan 23, 2021 at 16:07 | history | suggested | E. A. | CC BY-SA 4.0 |
I removed import statements that were not used in the migrations file. The author most likely copied an existing migration file that had needed those imports and replaced the contents without discarding unused import statements.
|
| Jan 22, 2021 at 22:52 | review | Suggested edits | |||
| S Jan 23, 2021 at 16:07 | |||||
| Jan 7, 2019 at 19:36 | comment | added | roOt |
I have used something similar where I combined rank and similarity like this: score=(F("rank") + F("similarity")) / 2
|
|
| S Nov 21, 2018 at 5:39 | history | suggested | wengole | CC BY-SA 4.0 |
I'm not 100% sure, but shouldn't `search` be `search_query` in the `similarity` annotation?
|
| Nov 20, 2018 at 14:48 | review | Suggested edits | |||
| S Nov 21, 2018 at 5:39 | |||||
| Sep 27, 2018 at 12:26 | comment | added | Thomas Frössman | you could also sort by the greatest of both rank/score values if they are comparable for your use case: ``` return queryset.annotate(rank=SearchRank(vector, search_query) similarity=TrigramSimilarity( 'name', search ) + TrigramSimilarity( 'content', search ), best_score=Greatest('rank', 'similarity') ).filter(Q(best_score__gte=0.3)).order_by('-best_score')[:20] ``` | |
| May 8, 2017 at 9:11 | comment | added | Private | Thanks for sharing the solution you found. | |
| Jul 11, 2016 at 15:07 | vote | accept | SalahAdDin | ||
| Jun 18, 2016 at 4:52 | history | answered | SalahAdDin | CC BY-SA 3.0 |