You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Using master, it looks like the @mapmany command is missing. For the groupby and then split approach mapmany is critical, is it just not implemented? I didn't see it mentioned in the codebase.
julia> db = SQLite.DB("data.sqlite3");
julia>getproperty(db, tbl[1]) |>@groupby(_.cell_id) |># filter stuff@map({cell_id=key(_), rows=_}) |>@mapmany(i->i.rows, (i, j)->{j...}) |>
DataFrame;
MethodError: no method matching mapmany(::QuerySQLite.SourceCode{SQLite.DB}, ::var"#113#123", ::Expr, ::var"#114#124", ::Expr)
Closest candidates are:mapmany(!Matched::QueryOperators.Enumerable, ::Function, ::Expr, ::Function, ::Expr) at /home/tlnagy/.julia/packages/QueryOperators/g4G21/src/enumerable/enumerable_mapmany.jl:42
Stacktrace:
[1] (::var"#112#122")(::QuerySQLite.SourceCode{SQLite.DB}) at /home/tlnagy/.julia/packages/Query/AwBtd/src/standalone_query_macros.jl:205
[2] |>(::QuerySQLite.SourceCode{SQLite.DB}, ::var"#112#122") at ./operators.jl:823
[3] top-level scope at In[30]:1
Hmm, I'm not sure that would do it...do you think you could come up with an example of using @mapmany using the Chinook_Sqlite.sqlite data in the test folder of the package (reading it into julia first, maybe as a dataframe). Then I'll see if I can replicate it in SQL?
Activity
bramtayl commentedon Jul 16, 2020
Nope, just not implemented, but shouldn't be too hard. Do you happen to know what the corresponding SQL code would look like?
tlnagy commentedon Jul 19, 2020
I'm pretty new to SQL, but maybe the UNIONALL command is what we're looking for? I'm not sure it covers all of
@mapmany's functionality.bramtayl commentedon Jul 20, 2020
Hmm, I'm not sure that would do it...do you think you could come up with an example of using
@mapmanyusing theChinook_Sqlite.sqlitedata in the test folder of the package (reading it into julia first, maybe as a dataframe). Then I'll see if I can replicate it in SQL?