| Safe Haskell | None |
|---|---|
| Language | Haskell2010 |
DataFrame.Operations.Aggregation
Synopsis
- aggregate :: [NamedExpr] -> GroupedDataFrame -> DataFrame
- computeRowHashes :: [Int] -> DataFrame -> Vector Int
- distinct :: DataFrame -> DataFrame
- selectIndices :: Vector Int -> DataFrame -> DataFrame
- hashKeyUnboxed :: Unbox a => MVector s Int -> Maybe Bitmap -> (Int -> a -> Int) -> Vector a -> ST s ()
- groupBy :: [Text] -> DataFrame -> GroupedDataFrame
- buildRowToGroup :: Int -> Vector Int -> Vector Int -> Vector Int
- changingPoints :: Vector (Int, Int) -> Vector Int
Documentation
aggregate :: [NamedExpr] -> GroupedDataFrame -> DataFrame Source #
Aggregate a grouped dataframe using the expressions given. All ungrouped columns will be dropped.
hashKeyUnboxed :: Unbox a => MVector s Int -> Maybe Bitmap -> (Int -> a -> Int) -> Vector a -> ST s () Source #
Fold a value-mix over an unboxed key column into the running hash vector,
respecting the null bitmap (a null slot mixes nullSalt instead of its
uninitialised stored value, so Nothing never matches a present Just x).
The non-nullable case is the original tight loop. INLINEd to specialise per call.
groupBy :: [Text] -> DataFrame -> GroupedDataFrame #
O(k * n) groups the dataframe by the given rows aggregating the remaining rows into vector that should be reduced later.