Support building SimpleDMatrix from Arrow data format (#7512)

* Integrate with Arrow C data API.
* Support Arrow dataset.
* Support Arrow table.

Co-authored-by: Xiaochang Wu <xiaochang.wu@intel.com>
Co-authored-by: Jiaming Yuan <jm.yuan@outlook.com>
Co-authored-by: Zhang Zhang <zhang.zhang@intel.com>
This commit is contained in:
Xiaochang Wu
2022-03-14 22:25:19 -07:00
committed by GitHub
parent 6b6849b001
commit 613ec36c5a
14 changed files with 732 additions and 10 deletions

View File

@@ -1000,6 +1000,8 @@ template DMatrix *
DMatrix::Create(data::IteratorAdapter<DataIterHandle, XGBCallbackDataIterNext,
XGBoostBatchCSR> *adapter,
float missing, int nthread, const std::string &cache_prefix);
template DMatrix* DMatrix::Create<data::RecordBatchesIterAdapter>(
data::RecordBatchesIterAdapter* adapter, float missing, int nthread, const std::string&);
SparsePage SparsePage::GetTranspose(int num_columns, int32_t n_threads) const {
SparsePage transpose;