Logo ROOT  
Reference Guide
RArrowDS.hxx
Go to the documentation of this file.
1/*************************************************************************
2 * Copyright (C) 1995-2021, Rene Brun and Fons Rademakers. *
3 * All rights reserved. *
4 * *
5 * For the licensing terms see $ROOTSYS/LICENSE. *
6 * For the list of contributors see $ROOTSYS/README/CREDITS. *
7 *************************************************************************/
8
9#ifndef ROOT_RARROWTDS
10#define ROOT_RARROWTDS
11
12#include "ROOT/RDataFrame.hxx"
13#include "ROOT/RDataSource.hxx"
14
15#include <memory>
16
17namespace arrow {
18class Table;
19}
20
21namespace ROOT {
22namespace Internal {
23namespace RDF {
24class TValueGetter;
25} // namespace RDF
26} // namespace Internal
27
28namespace RDF {
29
30class RArrowDS final : public RDataSource {
31private:
32 std::shared_ptr<arrow::Table> fTable;
33 std::vector<std::pair<ULong64_t, ULong64_t>> fEntryRanges;
34 std::vector<std::string> fColumnNames;
35 size_t fNSlots = 0U;
36
37 std::vector<std::pair<size_t, size_t>> fGetterIndex; // (columnId, visitorId)
38 std::vector<std::unique_ptr<ROOT::Internal::RDF::TValueGetter>> fValueGetters; // Visitors to be used to track and get entries. One per column.
39 std::vector<void *> GetColumnReadersImpl(std::string_view name, const std::type_info &type) override;
40
41public:
42 RArrowDS(std::shared_ptr<arrow::Table> table, std::vector<std::string> const &columns);
43 ~RArrowDS();
44 const std::vector<std::string> &GetColumnNames() const override;
45 std::vector<std::pair<ULong64_t, ULong64_t>> GetEntryRanges() override;
46 std::string GetTypeName(std::string_view colName) const override;
47 bool HasColumn(std::string_view colName) const override;
48 bool SetEntry(unsigned int slot, ULong64_t entry) override;
49 void InitSlot(unsigned int slot, ULong64_t firstEntry) override;
50 void SetNSlots(unsigned int nSlots) override;
51 void Initialise() override;
52 std::string GetLabel() override;
53};
54
55////////////////////////////////////////////////////////////////////////////////////////////////
56/// \brief Factory method to create a Apache Arrow RDataFrame.
57/// \param[in] table an apache::arrow table to use as a source.
58RDataFrame MakeArrowDataFrame(std::shared_ptr<arrow::Table> table, std::vector<std::string> const &columns);
59
60} // namespace RDF
61
62} // namespace ROOT
63
64#endif
unsigned long long ULong64_t
Definition: RtypesCore.h:81
char name[80]
Definition: TGX11.cxx:110
int type
Definition: TGX11.cxx:121
Helper class which keeps track for each slot where to get the entry.
Definition: RArrowDS.cxx:204
RDataFrame data source class to interface with Apache Arrow.
Definition: RArrowDS.hxx:30
bool HasColumn(std::string_view colName) const override
Checks if the dataset has a certain column.
Definition: RArrowDS.cxx:493
RArrowDS(std::shared_ptr< arrow::Table > table, std::vector< std::string > const &columns)
Constructor to create an Arrow RDataSource for RDataFrame.
Definition: RArrowDS.cxx:388
void Initialise() override
Convenience method called before starting an event-loop.
Definition: RArrowDS.cxx:589
std::string GetLabel() override
Return a string representation of the datasource type.
Definition: RArrowDS.cxx:595
void SetNSlots(unsigned int nSlots) override
Inform RDataSource of the number of processing slots (i.e.
Definition: RArrowDS.cxx:554
~RArrowDS()
Destructor.
Definition: RArrowDS.cxx:460
const std::vector< std::string > & GetColumnNames() const override
Returns a reference to the collection of the dataset's column names.
Definition: RArrowDS.cxx:464
void InitSlot(unsigned int slot, ULong64_t firstEntry) override
Convenience method called at the start of the data processing associated to a slot.
Definition: RArrowDS.cxx:511
std::vector< std::pair< ULong64_t, ULong64_t > > GetEntryRanges() override
Return ranges of entries to distribute to tasks.
Definition: RArrowDS.cxx:469
std::shared_ptr< arrow::Table > fTable
Definition: RArrowDS.hxx:32
std::vector< std::pair< size_t, size_t > > fGetterIndex
Definition: RArrowDS.hxx:37
std::vector< std::unique_ptr< ROOT::Internal::RDF::TValueGetter > > fValueGetters
Definition: RArrowDS.hxx:38
std::vector< void * > GetColumnReadersImpl(std::string_view name, const std::type_info &type) override
This needs to return a pointer to the pointer each value getter will point to.
Definition: RArrowDS.cxx:570
std::vector< std::string > fColumnNames
Definition: RArrowDS.hxx:34
std::string GetTypeName(std::string_view colName) const override
Type of a column as a string, e.g.
Definition: RArrowDS.cxx:475
std::vector< std::pair< ULong64_t, ULong64_t > > fEntryRanges
Definition: RArrowDS.hxx:33
bool SetEntry(unsigned int slot, ULong64_t entry) override
Advance the "cursors" returned by GetColumnReaders to the selected entry for a particular slot.
Definition: RArrowDS.cxx:502
RDataSource defines an API that RDataFrame can use to read arbitrary data formats.
ROOT's RDataFrame offers a high level interface for analyses of data stored in TTree,...
Definition: RDataFrame.hxx:40
basic_string_view< char > string_view
RDataFrame MakeArrowDataFrame(std::shared_ptr< arrow::Table > table, std::vector< std::string > const &columns)
Factory method to create a Apache Arrow RDataFrame.
Definition: RArrowDS.cxx:604
tbb::task_arena is an alias of tbb::interface7::task_arena, which doesn't allow to forward declare tb...