Trait ColumnarFrame 

pub trait ColumnarFrame<Row>: Sized {
    // Required method
    fn from_rows(rows: Vec<Row>) -> Self;

    // Provided method
    fn into_rows(self) -> Vec<Row>
       where Self: IntoIterator<Item = Row> { ... }
}

Rust rows ↔ the pure-Rust columnar companion generated by #[derive(DataFrameRow)].

#[derive(DataFrameRow)] builds a companion struct (<Row>DataFrame) whose fields are Vec-columns — the in-memory, column-oriented form of Vec<Row>. This trait is the documented verb surface for that companion: build it from rows (sequentially, or in parallel with feature = "rayon") and, for row-iterable companions, read the rows back.

It is the pure-Rust complement to IntoDataFrame / FromDataFrame, which cross the R boundary (Vec<Row> ↔ R data.frame). In particular from_rows_par (feature = "rayon") builds the pure-Rust companion in parallel — IntoDataFrame::into_dataframe_par cannot, since it yields an R-backed BuiltDataFrame rather than the companion.

The derive implements this on the companion type; call CompanionType::from_rows(rows). Row is a type parameter (not an associated type) so the derive can implement it for a pub companion even when the row type is private — mirroring From<Vec<Row>>, and avoiding the private-type-in-public-interface error an associated type would raise.

Reading rows back

into_rows is available for row-iterable companions (named-field structs without column expansion), where it defers to the companion’s IntoIterator. Enum and column-expansion companions are write-only in pure Rust; read those back across the R boundary via FromDataFrame or the derive’s one-call try_from_dataframe.

Required Methods

fn from_rows(rows: Vec<Row>) -> Self

Transpose a row vector into the columnar companion (sequential).

Provided Methods

fn into_rows(self) -> Vec<Row>

where Self: IntoIterator<Item = Row>,

Read the rows back out of the companion.

Available for row-iterable companions (named-field structs without column expansion); defers to the companion’s IntoIterator. Enum and column-expansion companions are write-only in pure Rust — read those back via FromDataFrame / try_from_dataframe.

Dyn Compatibility

This trait is not dyn compatible.

In older versions of Rust, dyn compatibility was called "object safety".

Implementors