from __future__ import annotations import operator from typing import TYPE_CHECKING from typing import Any from typing import Callable from typing import Literal from typing import Sequence from typing import cast from narwhals._compliant import LazyExpr from narwhals._expression_parsing import ExprKind from narwhals._spark_like.expr_dt import SparkLikeExprDateTimeNamespace from narwhals._spark_like.expr_list import SparkLikeExprListNamespace from narwhals._spark_like.expr_str import SparkLikeExprStringNamespace from narwhals._spark_like.expr_struct import SparkLikeExprStructNamespace from narwhals._spark_like.utils import WindowInputs from narwhals._spark_like.utils import import_functions from narwhals._spark_like.utils import import_native_dtypes from narwhals._spark_like.utils import import_window from narwhals._spark_like.utils import narwhals_to_native_dtype from narwhals.dependencies import get_pyspark from narwhals.utils import Implementation from narwhals.utils import not_implemented from narwhals.utils import parse_version if TYPE_CHECKING: from sqlframe.base.column import Column from sqlframe.base.window import Window from typing_extensions import Self from narwhals._compliant.typing import AliasNames from narwhals._compliant.typing import EvalNames from narwhals._compliant.typing import EvalSeries from narwhals._expression_parsing import ExprMetadata from narwhals._spark_like.dataframe import SparkLikeLazyFrame from narwhals._spark_like.namespace import SparkLikeNamespace from narwhals._spark_like.typing import WindowFunction from narwhals.dtypes import DType from narwhals.typing import FillNullStrategy from narwhals.typing import NonNestedLiteral from narwhals.typing import NumericLiteral from narwhals.typing import RankMethod from narwhals.typing import TemporalLiteral from narwhals.utils import Version from narwhals.utils import _FullContext class SparkLikeExpr(LazyExpr["SparkLikeLazyFrame", "Column"]): def __init__( self, call: EvalSeries[SparkLikeLazyFrame, Column], *, evaluate_output_names: EvalNames[SparkLikeLazyFrame], alias_output_names: AliasNames | None, backend_version: tuple[int, ...], version: Version, implementation: Implementation, ) -> None: self._call = call self._evaluate_output_names = evaluate_output_names self._alias_output_names = alias_output_names self._backend_version = backend_version self._version = version self._implementation = implementation self._window_function: WindowFunction | None = None self._metadata: ExprMetadata | None = None def __call__(self, df: SparkLikeLazyFrame) -> Sequence[Column]: return self._call(df) def broadcast(self, kind: Literal[ExprKind.AGGREGATION, ExprKind.LITERAL]) -> Self: if kind is ExprKind.LITERAL: return self def func(df: SparkLikeLazyFrame) -> Sequence[Column]: return [ result.over(self._Window().partitionBy(self._F.lit(1))) for result in self(df) ] return self.__class__( func, evaluate_output_names=self._evaluate_output_names, alias_output_names=self._alias_output_names, backend_version=self._backend_version, version=self._version, implementation=self._implementation, ) @property def _F(self): # type: ignore[no-untyped-def] # noqa: ANN202, N802 if TYPE_CHECKING: from sqlframe.base import functions return functions else: return import_functions(self._implementation) @property def _native_dtypes(self): # type: ignore[no-untyped-def] # noqa: ANN202 if TYPE_CHECKING: from sqlframe.base import types return types else: return import_native_dtypes(self._implementation) @property def _Window(self) -> type[Window]: # noqa: N802 if TYPE_CHECKING: from sqlframe.base.window import Window return Window else: return import_window(self._implementation) def __narwhals_expr__(self) -> None: ... def __narwhals_namespace__(self) -> SparkLikeNamespace: # pragma: no cover # Unused, just for compatibility with PandasLikeExpr from narwhals._spark_like.namespace import SparkLikeNamespace return SparkLikeNamespace( backend_version=self._backend_version, version=self._version, implementation=self._implementation, ) def _with_window_function(self, window_function: WindowFunction) -> Self: result = self.__class__( self._call, evaluate_output_names=self._evaluate_output_names, alias_output_names=self._alias_output_names, backend_version=self._backend_version, version=self._version, implementation=self._implementation, ) result._window_function = window_function return result def _cum_window_func( self, *, reverse: bool, func_name: Literal["sum", "max", "min", "count", "product"], ) -> WindowFunction: def func(window_inputs: WindowInputs) -> Column: if reverse: order_by_cols = [ self._F.col(x).desc_nulls_last() for x in window_inputs.order_by ] else: order_by_cols = [ self._F.col(x).asc_nulls_first() for x in window_inputs.order_by ] window = ( self._Window() .partitionBy(list(window_inputs.partition_by)) .orderBy(order_by_cols) .rowsBetween(self._Window().unboundedPreceding, 0) ) return getattr(self._F, func_name)(window_inputs.expr).over(window) return func def _rolling_window_func( self, *, func_name: Literal["sum", "mean", "std", "var"], center: bool, window_size: int, min_samples: int, ddof: int | None = None, ) -> WindowFunction: supported_funcs = ["sum", "mean", "std", "var"] if center: half = (window_size - 1) // 2 remainder = (window_size - 1) % 2 start = self._Window().currentRow - half - remainder end = self._Window().currentRow + half else: start = self._Window().currentRow - window_size + 1 end = self._Window().currentRow def func(window_inputs: WindowInputs) -> Column: window = ( self._Window() .partitionBy(list(window_inputs.partition_by)) .orderBy( [self._F.col(x).asc_nulls_first() for x in window_inputs.order_by] ) .rowsBetween(start, end) ) if func_name in {"sum", "mean"}: func_: str = func_name elif func_name == "var" and ddof == 0: func_ = "var_pop" elif func_name in "var" and ddof == 1: func_ = "var_samp" elif func_name == "std" and ddof == 0: func_ = "stddev_pop" elif func_name == "std" and ddof == 1: func_ = "stddev_samp" elif func_name in {"var", "std"}: # pragma: no cover msg = f"Only ddof=0 and ddof=1 are currently supported for rolling_{func_name}." raise ValueError(msg) else: # pragma: no cover msg = f"Only the following functions are supported: {supported_funcs}.\nGot: {func_name}." raise ValueError(msg) return self._F.when( self._F.count(window_inputs.expr).over(window) >= min_samples, getattr(self._F, func_)(window_inputs.expr).over(window), ) return func @classmethod def from_column_names( cls: type[Self], evaluate_column_names: EvalNames[SparkLikeLazyFrame], /, *, context: _FullContext, ) -> Self: def func(df: SparkLikeLazyFrame) -> list[Column]: return [df._F.col(col_name) for col_name in evaluate_column_names(df)] return cls( func, evaluate_output_names=evaluate_column_names, alias_output_names=None, backend_version=context._backend_version, version=context._version, implementation=context._implementation, ) @classmethod def from_column_indices( cls: type[Self], *column_indices: int, context: _FullContext ) -> Self: def func(df: SparkLikeLazyFrame) -> list[Column]: columns = df.columns return [df._F.col(columns[i]) for i in column_indices] return cls( func, evaluate_output_names=lambda df: [df.columns[i] for i in column_indices], alias_output_names=None, backend_version=context._backend_version, version=context._version, implementation=context._implementation, ) def _with_callable( self, call: Callable[..., Column], /, **expressifiable_args: Self | Any, ) -> Self: def func(df: SparkLikeLazyFrame) -> list[Column]: native_series_list = self(df) lit = df._F.lit other_native_series = { key: df._evaluate_expr(value) if self._is_expr(value) else lit(value) for key, value in expressifiable_args.items() } return [ call(native_series, **other_native_series) for native_series in native_series_list ] return self.__class__( func, evaluate_output_names=self._evaluate_output_names, alias_output_names=self._alias_output_names, backend_version=self._backend_version, version=self._version, implementation=self._implementation, ) def _with_alias_output_names(self, func: AliasNames | None, /) -> Self: return type(self)( call=self._call, evaluate_output_names=self._evaluate_output_names, alias_output_names=func, backend_version=self._backend_version, version=self._version, implementation=self._implementation, ) def __eq__(self, other: SparkLikeExpr) -> Self: # type: ignore[override] return self._with_callable( lambda _input, other: _input.__eq__(other), other=other ) def __ne__(self, other: SparkLikeExpr) -> Self: # type: ignore[override] return self._with_callable( lambda _input, other: _input.__ne__(other), other=other ) def __add__(self, other: SparkLikeExpr) -> Self: return self._with_callable( lambda _input, other: _input.__add__(other), other=other ) def __sub__(self, other: SparkLikeExpr) -> Self: return self._with_callable( lambda _input, other: _input.__sub__(other), other=other ) def __rsub__(self, other: SparkLikeExpr) -> Self: return self._with_callable( lambda _input, other: other.__sub__(_input), other=other ).alias("literal") def __mul__(self, other: SparkLikeExpr) -> Self: return self._with_callable( lambda _input, other: _input.__mul__(other), other=other ) def __truediv__(self, other: SparkLikeExpr) -> Self: return self._with_callable( lambda _input, other: _input.__truediv__(other), other=other ) def __rtruediv__(self, other: SparkLikeExpr) -> Self: return self._with_callable( lambda _input, other: other.__truediv__(_input), other=other ).alias("literal") def __floordiv__(self, other: SparkLikeExpr) -> Self: def _floordiv(_input: Column, other: Column) -> Column: return self._F.floor(_input / other) return self._with_callable(_floordiv, other=other) def __rfloordiv__(self, other: SparkLikeExpr) -> Self: def _rfloordiv(_input: Column, other: Column) -> Column: return self._F.floor(other / _input) return self._with_callable(_rfloordiv, other=other).alias("literal") def __pow__(self, other: SparkLikeExpr) -> Self: return self._with_callable( lambda _input, other: _input.__pow__(other), other=other ) def __rpow__(self, other: SparkLikeExpr) -> Self: return self._with_callable( lambda _input, other: other.__pow__(_input), other=other ).alias("literal") def __mod__(self, other: SparkLikeExpr) -> Self: return self._with_callable( lambda _input, other: _input.__mod__(other), other=other ) def __rmod__(self, other: SparkLikeExpr) -> Self: return self._with_callable( lambda _input, other: other.__mod__(_input), other=other ).alias("literal") def __ge__(self, other: SparkLikeExpr) -> Self: return self._with_callable( lambda _input, other: _input.__ge__(other), other=other ) def __gt__(self, other: SparkLikeExpr) -> Self: return self._with_callable(lambda _input, other: _input > other, other=other) def __le__(self, other: SparkLikeExpr) -> Self: return self._with_callable( lambda _input, other: _input.__le__(other), other=other ) def __lt__(self, other: SparkLikeExpr) -> Self: return self._with_callable( lambda _input, other: _input.__lt__(other), other=other ) def __and__(self, other: SparkLikeExpr) -> Self: return self._with_callable( lambda _input, other: _input.__and__(other), other=other ) def __or__(self, other: SparkLikeExpr) -> Self: return self._with_callable( lambda _input, other: _input.__or__(other), other=other ) def __invert__(self) -> Self: invert = cast("Callable[..., Column]", operator.invert) return self._with_callable(invert) def abs(self) -> Self: return self._with_callable(self._F.abs) def all(self) -> Self: return self._with_callable(self._F.bool_and) def any(self) -> Self: return self._with_callable(self._F.bool_or) def cast(self, dtype: DType | type[DType]) -> Self: def _cast(_input: Column) -> Column: spark_dtype = narwhals_to_native_dtype( dtype, self._version, self._native_dtypes ) return _input.cast(spark_dtype) return self._with_callable(_cast) def count(self) -> Self: return self._with_callable(self._F.count) def max(self) -> Self: return self._with_callable(self._F.max) def mean(self) -> Self: return self._with_callable(self._F.mean) def median(self) -> Self: def _median(_input: Column) -> Column: if ( self._implementation in {Implementation.PYSPARK, Implementation.PYSPARK_CONNECT} and (pyspark := get_pyspark()) is not None and parse_version(pyspark) < (3, 4) ): # pragma: no cover # Use percentile_approx with default accuracy parameter (10000) return self._F.percentile_approx(_input.cast("double"), 0.5) return self._F.median(_input) return self._with_callable(_median) def min(self) -> Self: return self._with_callable(self._F.min) def null_count(self) -> Self: def _null_count(_input: Column) -> Column: return self._F.count_if(self._F.isnull(_input)) return self._with_callable(_null_count) def sum(self) -> Self: return self._with_callable(self._F.sum) def std(self, ddof: int) -> Self: from functools import partial import numpy as np # ignore-banned-import from narwhals._spark_like.utils import _std func = partial( _std, ddof=ddof, np_version=parse_version(np), functions=self._F, implementation=self._implementation, ) return self._with_callable(func) def var(self, ddof: int) -> Self: from functools import partial import numpy as np # ignore-banned-import from narwhals._spark_like.utils import _var func = partial( _var, ddof=ddof, np_version=parse_version(np), functions=self._F, implementation=self._implementation, ) return self._with_callable(func) def clip( self, lower_bound: Self | NumericLiteral | TemporalLiteral | None = None, upper_bound: Self | NumericLiteral | TemporalLiteral | None = None, ) -> Self: def _clip_lower(_input: Column, lower_bound: Column) -> Column: result = _input return self._F.when(result < lower_bound, lower_bound).otherwise(result) def _clip_upper(_input: Column, upper_bound: Column) -> Column: result = _input return self._F.when(result > upper_bound, upper_bound).otherwise(result) def _clip_both( _input: Column, lower_bound: Column, upper_bound: Column ) -> Column: return ( self._F.when(_input < lower_bound, lower_bound) .when(_input > upper_bound, upper_bound) .otherwise(_input) ) if lower_bound is None: return self._with_callable(_clip_upper, upper_bound=upper_bound) if upper_bound is None: return self._with_callable(_clip_lower, lower_bound=lower_bound) return self._with_callable( _clip_both, lower_bound=lower_bound, upper_bound=upper_bound ) def is_finite(self) -> Self: def _is_finite(_input: Column) -> Column: # A value is finite if it's not NaN, and not infinite, while NULLs should be # preserved is_finite_condition = ( ~self._F.isnan(_input) & (_input != self._F.lit(float("inf"))) & (_input != self._F.lit(float("-inf"))) ) return self._F.when(~self._F.isnull(_input), is_finite_condition).otherwise( None ) return self._with_callable(_is_finite) def is_in(self, values: Sequence[Any]) -> Self: def _is_in(_input: Column) -> Column: return _input.isin(values) if values else self._F.lit(False) # noqa: FBT003 return self._with_callable(_is_in) def is_unique(self) -> Self: def _is_unique(_input: Column) -> Column: # Create a window spec that treats each value separately return self._F.count("*").over(self._Window.partitionBy(_input)) == 1 return self._with_callable(_is_unique) def len(self) -> Self: def _len(_input: Column) -> Column: # Use count(*) to count all rows including nulls return self._F.count("*") return self._with_callable(_len) def round(self, decimals: int) -> Self: def _round(_input: Column) -> Column: return self._F.round(_input, decimals) return self._with_callable(_round) def skew(self) -> Self: return self._with_callable(self._F.skewness) def n_unique(self) -> Self: def _n_unique(_input: Column) -> Column: return self._F.count_distinct(_input) + self._F.max( self._F.isnull(_input).cast(self._native_dtypes.IntegerType()) ) return self._with_callable(_n_unique) def over(self, partition_by: Sequence[str], order_by: Sequence[str] | None) -> Self: if (window_function := self._window_function) is not None: assert order_by is not None # noqa: S101 def func(df: SparkLikeLazyFrame) -> list[Column]: return [ window_function( WindowInputs(expr, partition_by or [self._F.lit(1)], order_by) ) for expr in self._call(df) ] else: def func(df: SparkLikeLazyFrame) -> list[Column]: return [ expr.over(self._Window.partitionBy(*partition_by)) for expr in self._call(df) ] return self.__class__( func, evaluate_output_names=self._evaluate_output_names, alias_output_names=self._alias_output_names, backend_version=self._backend_version, version=self._version, implementation=self._implementation, ) def is_null(self) -> Self: return self._with_callable(self._F.isnull) def is_nan(self) -> Self: def _is_nan(_input: Column) -> Column: return self._F.when(self._F.isnull(_input), None).otherwise( self._F.isnan(_input) ) return self._with_callable(_is_nan) def shift(self, n: int) -> Self: def func(window_inputs: WindowInputs) -> Column: order_by_cols = [ self._F.col(x).asc_nulls_first() for x in window_inputs.order_by ] window = ( self._Window() .partitionBy(list(window_inputs.partition_by)) .orderBy(order_by_cols) ) return self._F.lag(window_inputs.expr, n).over(window) return self._with_window_function(func) def is_first_distinct(self) -> Self: def func(window_inputs: WindowInputs) -> Column: order_by_cols = [ self._F.col(x).asc_nulls_first() for x in window_inputs.order_by ] window = ( self._Window() .partitionBy([*window_inputs.partition_by, window_inputs.expr]) .orderBy(order_by_cols) ) return self._F.row_number().over(window) == 1 return self._with_window_function(func) def is_last_distinct(self) -> Self: def func(window_inputs: WindowInputs) -> Column: order_by_cols = [ self._F.col(x).desc_nulls_last() for x in window_inputs.order_by ] window = ( self._Window() .partitionBy([*window_inputs.partition_by, window_inputs.expr]) .orderBy(order_by_cols) ) return self._F.row_number().over(window) == 1 return self._with_window_function(func) def diff(self) -> Self: def func(window_inputs: WindowInputs) -> Column: order_by_cols = [ self._F.col(x).asc_nulls_first() for x in window_inputs.order_by ] window = ( self._Window() .partitionBy(list(window_inputs.partition_by)) .orderBy(order_by_cols) ) return window_inputs.expr - self._F.lag(window_inputs.expr).over(window) return self._with_window_function(func) def cum_sum(self, *, reverse: bool) -> Self: return self._with_window_function( self._cum_window_func(reverse=reverse, func_name="sum") ) def cum_max(self, *, reverse: bool) -> Self: return self._with_window_function( self._cum_window_func(reverse=reverse, func_name="max") ) def cum_min(self, *, reverse: bool) -> Self: return self._with_window_function( self._cum_window_func(reverse=reverse, func_name="min") ) def cum_count(self, *, reverse: bool) -> Self: return self._with_window_function( self._cum_window_func(reverse=reverse, func_name="count") ) def cum_prod(self, *, reverse: bool) -> Self: return self._with_window_function( self._cum_window_func(reverse=reverse, func_name="product") ) def fill_null( self, value: Self | NonNestedLiteral, strategy: FillNullStrategy | None, limit: int | None, ) -> Self: if strategy is not None: def _fill_with_strategy(window_inputs: WindowInputs) -> Column: fill_func = ( self._F.last_value if strategy == "forward" else self._F.first_value ) if strategy == "forward": start = ( -limit if limit is not None else self._Window().unboundedPreceding ) end = self._Window().currentRow else: start = self._Window().currentRow end = ( limit if limit is not None else self._Window().unboundedFollowing ) window = ( self._Window() .partitionBy(list(window_inputs.partition_by) or self._F.lit(1)) .orderBy( [self._F.col(x).asc_nulls_first() for x in window_inputs.order_by] ) .rowsBetween(start, end) ) return fill_func(window_inputs.expr, ignoreNulls=True).over(window) return self._with_window_function(_fill_with_strategy) def _fill_constant(_input: Column, value: Column) -> Column: return self._F.ifnull(_input, value) return self._with_callable(_fill_constant, value=value) def rolling_sum(self, window_size: int, *, min_samples: int, center: bool) -> Self: return self._with_window_function( self._rolling_window_func( func_name="sum", center=center, window_size=window_size, min_samples=min_samples, ) ) def rolling_mean(self, window_size: int, *, min_samples: int, center: bool) -> Self: return self._with_window_function( self._rolling_window_func( func_name="mean", center=center, window_size=window_size, min_samples=min_samples, ) ) def rolling_var( self, window_size: int, *, min_samples: int, center: bool, ddof: int ) -> Self: return self._with_window_function( self._rolling_window_func( func_name="var", center=center, window_size=window_size, min_samples=min_samples, ddof=ddof, ) ) def rolling_std( self, window_size: int, *, min_samples: int, center: bool, ddof: int ) -> Self: return self._with_window_function( self._rolling_window_func( func_name="std", center=center, window_size=window_size, min_samples=min_samples, ddof=ddof, ) ) def rank(self, method: RankMethod, *, descending: bool) -> Self: if method in {"min", "max", "average"}: func_name = "rank" elif method == "dense": func_name = "dense_rank" else: # method == "ordinal" func_name = "row_number" def _rank(_input: Column) -> Column: if descending: order_by_cols = [self._F.desc_nulls_last(_input)] else: order_by_cols = [self._F.asc_nulls_last(_input)] window = self._Window().partitionBy(self._F.lit(1)).orderBy(order_by_cols) count_window = self._Window().partitionBy(_input) if method == "max": expr = ( getattr(self._F, func_name)().over(window) + self._F.count(_input).over(count_window) - self._F.lit(1) ) elif method == "average": expr = getattr(self._F, func_name)().over(window) + ( self._F.count(_input).over(count_window) - self._F.lit(1) ) / self._F.lit(2) else: expr = getattr(self._F, func_name)().over(window) return self._F.when(_input.isNotNull(), expr) return self._with_callable(_rank) @property def str(self) -> SparkLikeExprStringNamespace: return SparkLikeExprStringNamespace(self) @property def dt(self) -> SparkLikeExprDateTimeNamespace: return SparkLikeExprDateTimeNamespace(self) @property def list(self) -> SparkLikeExprListNamespace: return SparkLikeExprListNamespace(self) @property def struct(self) -> SparkLikeExprStructNamespace: return SparkLikeExprStructNamespace(self) drop_nulls = not_implemented() unique = not_implemented() quantile = not_implemented()