| Safe Haskell | None |
|---|---|
| Language | Haskell2010 |
DataFrame.IO.Parquet.Dictionary
Synopsis
- dictCardinality :: DictVals -> Int
- readDictVals :: ParquetType -> ByteString -> Maybe Int32 -> DictVals
- readPageInt32 :: ByteString -> [Int32]
- readPageWord64 :: ByteString -> [Double]
- readPageBytes :: ByteString -> [Text]
- readPageBool :: ByteString -> [Bool]
- readPageInt64 :: ByteString -> [Int64]
- readPageFloat :: ByteString -> [Float]
- readNInt96Times :: Int -> ByteString -> ([UTCTime], ByteString)
- readPageInt96Times :: ByteString -> [UTCTime]
- readPageFixedBytes :: ByteString -> Int -> [Text]
- stitchForRepBool :: Int -> Int -> [Int] -> [Int] -> [Bool] -> Column
- stitchForRepInt32 :: Int -> Int -> [Int] -> [Int] -> [Int32] -> Column
- stitchForRepInt64 :: Int -> Int -> [Int] -> [Int] -> [Int64] -> Column
- stitchForRepUTCTime :: Int -> Int -> [Int] -> [Int] -> [UTCTime] -> Column
- stitchForRepFloat :: Int -> Int -> [Int] -> [Int] -> [Float] -> Column
- stitchForRepDouble :: Int -> Int -> [Int] -> [Int] -> [Double] -> Column
- stitchForRepText :: Int -> Int -> [Int] -> [Int] -> [Text] -> Column
- applyDictToColumn :: (Columnable a, Columnable (Maybe a)) => Vector a -> Vector Int -> Int -> [Int] -> IO Column
- decodeDictV1 :: Maybe DictVals -> Int -> Int -> [Int] -> [Int] -> Int -> ByteString -> IO Column
- toMaybeInt32 :: Int -> [Int] -> [Int32] -> Column
- toMaybeDouble :: Int -> [Int] -> [Double] -> Column
- toMaybeText :: Int -> [Int] -> [Text] -> Column
- toMaybeBool :: Int -> [Int] -> [Bool] -> Column
- toMaybeInt64 :: Int -> [Int] -> [Int64] -> Column
- toMaybeFloat :: Int -> [Int] -> [Float] -> Column
- toMaybeUTCTime :: Int -> [Int] -> [UTCTime] -> Column
Documentation
dictCardinality :: DictVals -> Int Source #
readDictVals :: ParquetType -> ByteString -> Maybe Int32 -> DictVals Source #
readPageInt32 :: ByteString -> [Int32] Source #
readPageWord64 :: ByteString -> [Double] Source #
readPageBytes :: ByteString -> [Text] Source #
readPageBool :: ByteString -> [Bool] Source #
readPageInt64 :: ByteString -> [Int64] Source #
readPageFloat :: ByteString -> [Float] Source #
readNInt96Times :: Int -> ByteString -> ([UTCTime], ByteString) Source #
readPageInt96Times :: ByteString -> [UTCTime] Source #
readPageFixedBytes :: ByteString -> Int -> [Text] Source #
stitchForRepBool :: Int -> Int -> [Int] -> [Int] -> [Bool] -> Column Source #
Dispatch to the right multi-level list stitching function. For maxRep=1 uses stitchList; for 23 uses stitchList23 with computed thresholds. Threshold formula: defT_r = maxDef - 2*(maxRep - r).
applyDictToColumn :: (Columnable a, Columnable (Maybe a)) => Vector a -> Vector Int -> Int -> [Int] -> IO Column Source #
Build a Column from a dictionary + index vector + def levels in a single mutable-vector pass, avoiding the intermediate [a] and [Maybe a] lists. For maxRep > 0 (list columns) the caller must use the rep-stitching path instead.