Hey guys,
given the following data, what I’d like to do is do a GROUPBY with REDUCE TOLIST, but also filter on the LIST items prior to returning the result.
From this result set, I would only like to return the results that have a Type of 2 in the grouping, so in this case the first result would not be returned because BookID of 2 only has a Type of 1.
Hey Mark,
that is close, I need @Type==2, but in the list of results, I would like to also return all of the other types, so in my example above the result returned should be:
(integer) 2
“Title”
“book1”
“BookID”
“1” - type 1 but keep it since the one below is of Type 2
Oh, you wish to return all the IDs of the title, if any of those books are of type 2. This would require an aggregation functionality which would (1) create a set of all types in the reduced entries, and (2) filter through this list, omitting any which don’t have type 1 as a member. Currently such a function doesn’t exist, but I guess it could be implemented if the use case were compelling enough. This does sound more like a use case for redis gears, though.
Mark Nunberg | Senior Software Engineer Redis Labs - home of Redis
Quite a puzzle. I think I captured what you’re trying to do.
I was able to accomplish this by doing a query like this:
“FT.AGGREGATE” “books_hash_index_1” “*” LOAD 3 @Title@Type@BookID APPLY “100^(@type)” As TypeNumber GROUPBY 1 @Title REDUCE TOLIST 1 @BookID AS BookID REDUCE SUM 1 @TypeNumber as SumType APPLY “@SumType >= 1000” as Type2 APPLY “(@SumType >= 100) && (@SumType < 1000)” As Type1 FILTER “@Type2 == 1”
The results look like this:
(integer) 3
“Title”
“book1”
“BookID”
“1”
“3”
“SumType”
“10100”
“Type2”
“1”
“Type1”
“0”
“Title”
“book3”
“BookID”
“4”
“SumType”
“10000”
“Type2”
“1”
“Type1”
“0”
I know this is a lot, but it’s really not that complicated. What I’m doing is creating a pseudo-bitmask out of the type field by using exponents. The later, I’m using some logic to check if it is a member of a particular type. You don’t really need to create fields, but I’m doing it for clarity. Finally, I’m filtering on the a particular type.
Hey Kyle,
that does look like the desired result! The only problem I notice is the count is 3, instead of 2, do you think it’s counting the BookIDs instead of the result set?
If I could get this one last bit correct, I would be all set!
