Struct LazyFrame 

pub struct LazyFrame { /* private fields */ }

Lazy query API (LazyFrame, LazyGroupBy). A lazily-evaluated query backed by a LogicalPlan.

Implementations

impl LazyFrame

pub fn from_dataframe(df: DataFrame) -> Self

Create a LazyFrame that scans an in-memory DataFrame.

pub fn scan_csv(path: impl AsRef<Path>) -> Result<Self>

Build a CSV scan plan (no file I/O is performed until collect()).

pub fn scan_parquet(path: impl AsRef<Path>) -> Result<Self>

Build a Parquet scan plan (no file I/O is performed until collect()).

pub fn concat(inputs: Vec<LazyFrame>) -> Result<Self>

Construct strict vertical concatenation from two or more lazy inputs.

When every schema is already known, mismatch is reported at plan construction. For lazy file scans whose schema is intentionally unavailable until bounded open, the streaming executor preflights every child before its first output and rejects any mismatch then.

pub fn select(self, exprs: Vec<Expr>) -> Self

Add a projection (select) node to the logical plan.

pub fn filter(self, predicate: Expr) -> Self

Add a filter node to the logical plan.

pub fn with_columns(self, exprs: Vec<Expr>) -> Self

Add a projection (with_columns) node to the logical plan.

pub fn group_by(self, by: Vec<Expr>) -> LazyGroupBy

Start a group-by on this LazyFrame.

pub fn join<K: Into<JoinKeys>>( self, other: LazyFrame, keys: K, how: JoinType, ) -> Self

Join with another LazyFrame using provided join keys.

pub fn sort(self, options: SortOptions) -> Self

Sort by one or more columns.

pub fn head(self, n: usize) -> Self

Return the first n rows.

pub fn tail(self, n: usize) -> Self

Return the last n rows.

pub fn unique(self, subset: Option<Vec<String>>) -> Self

Remove duplicate rows.

pub fn fill_null<T: Into<FillNull>>(self, fill: T) -> Self

Fill null values using a scalar or strategy.

pub fn drop_nulls(self, subset: Option<Vec<String>>) -> Self

Drop rows containing null values.

pub fn null_count(self) -> Self

Count null values per column.

pub fn explode(self, column: impl Into<String>) -> Self

Explode one List<Utf8> column into multiple rows.

pub fn implode(self) -> Self

Implode UTF-8 columns into one row of List<Utf8> columns.

pub fn collect(self) -> Result<DataFrame>

Optimize, compile, and execute this LazyFrame into an eager DataFrame.

pub fn collect_with_options(self, options: StreamOptions) -> Result<DataFrame>

Materialize a supported bounded stream while reserving every retained output batch.

This deliberately uses the streaming source/operator path and never wraps the eager executor with accounting after allocation.

pub fn collect_streaming( self, options: StreamOptions, ) -> Result<DataFrameStream>

Compile this plan for incremental bounded execution.

The compiler performs eligibility analysis before any source is opened. Until a source’s v0.8 BatchSourceFactory is installed, it returns StreamingUnsupported rather than calling the legacy eager executor.

pub fn explain(self, optimized: bool) -> String

Render the logical plan as a human-readable string.

If optimized is true, includes optimizer rewrites such as pushdowns.

Trait Implementations

impl Clone for LazyFrame

fn clone(&self) -> LazyFrame

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for LazyFrame

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.