LazyCSVLoader

LazyCSVLoader(csv_path, columns=None)

Lazy CSV loader using Polars scan_csv.

Per ADR-006: - Uses polars.scan_csv() for unevaluated query plan - .materialize() calls .collect() to return DataFrame - Falls back to pandas DataFrame when materialized

Attributes

Name	Description
columns	Column names. Creates lazy plan but does not load data.
shape	Shape estimate. Creates lazy plan but does not load data.

Methods

Name	Description
filter	Add lazy filter condition.
head	Add lazy head operation.
materialize	Force full CSV load and return as pandas DataFrame.
select	Add lazy column selection.
tail	Add lazy tail operation.
with_columns	Add lazy column computation.

filter

LazyCSVLoader.filter(*args, **kwargs)

Add lazy filter condition.

head

LazyCSVLoader.head(n=10)

Add lazy head operation.

materialize

LazyCSVLoader.materialize()

Force full CSV load and return as pandas DataFrame.

select

LazyCSVLoader.select(*args)

Add lazy column selection.

tail

LazyCSVLoader.tail(n=10)

Add lazy tail operation.

with_columns

LazyCSVLoader.with_columns(*args, **kwargs)

Add lazy column computation.