Where one of the loader's columns lives inside a batch-tensor row and how it's shaped.
A batch row is a flat span of floats in which a vector column occupies fWidth consecutive slots (padded to its specified maximum size). This describes the inverse mapping, so a sink can write each column back under its original name (unexpanded).
Definition at line 38 of file RBatchSink.hxx.
Public Attributes | |
| bool | fIsVector |
| std::string | fName |
| std::size_t | fOffset |
| std::size_t | fWidth |
#include <ROOT/ML/RBatchSink.hxx>
| bool ROOT::Experimental::Internal::ML::RColumnLayout::fIsVector |
Definition at line 42 of file RBatchSink.hxx.
| std::string ROOT::Experimental::Internal::ML::RColumnLayout::fName |
Definition at line 39 of file RBatchSink.hxx.
| std::size_t ROOT::Experimental::Internal::ML::RColumnLayout::fOffset |
Definition at line 40 of file RBatchSink.hxx.
| std::size_t ROOT::Experimental::Internal::ML::RColumnLayout::fWidth |
Definition at line 41 of file RBatchSink.hxx.