# xreducer¶

Defined in xtensor/xreducer.hpp

template<class F, class CT, class X, class O>
class xreducer : public xt::xsharable_expression<xreducer<F, CT, X, O>>, public xt::xconst_iterable<xreducer<F, CT, X, O>>, public xt::xaccessible<xreducer<F, CT, X, O>>, public extension::xreducer_base_t<F, CT, X, O>

Reducing function operating over specified axes.

The xreducer class implements an xexpression applying a reducing function to an xexpression over the specified axes.

The reducer’s result_type is deduced from the result type of function

F::reduce_functor_type when called with elements of the expression
See
Template Parameters
• F: a tuple of functors (class xreducer_functors or compatible)

• CT: the closure type of the xexpression to reduce

• X: the list of axes

Template Parameters
• CT.:

Constructor

template<class Func, class CTA, class AX, class OX>
xreducer(Func &&func, CTA &&e, AX &&axes, OX &&options)

Constructs an xreducer expression applying the specified function to the given expression over the given axes.

Parameters
• func: the function to apply

• e: the expression to reduce

• axes: the axes along which the reduction is performed

Size and shape

auto shape() const

Returns the shape of the expression.

layout_type layout() const

Returns the shape of the expression.

Data

auto expression() const

Returns a constant reference to the underlying expression of the reducer.

template<class ...Args>
auto operator()(Args... args) const

Returns a constant reference to the element at the specified position in the reducer.

Parameters
• args: a list of indices specifying the position in the reducer. Indices must be unsigned integers, the number of indices should be equal or greater than the number of dimensions of the reducer.

template<class ...Args>
auto unchecked(Args... args) const

Returns a constant reference to the element at the specified position in the reducer.

Warning

This method is meant for performance, for expressions with a dynamic number of dimensions (i.e. not known at compile time). Since it may have undefined behavior (see parameters), operator() should be prefered whenever it is possible.

Warning

This method is NOT compatible with broadcasting, meaning the following code has undefined behavior:

xt::xarray<double> a = {{0, 1}, {2, 3}};
xt::xarray<double> b = {0, 1};
auto fd = a + b;
double res = fd.uncheked(0, 1);


Parameters
• args: a list of indices specifying the position in the reducer. Indices must be unsigned integers, the number of indices must be equal to the number of dimensions of the reducer, else the behavior is undefined.

template<class It>
auto element(It first, It last) const

Returns a constant reference to the element at the specified position in the reducer.

Parameters
• first: iterator starting the sequence of indices

• last: iterator ending the sequence of indices The number of indices in the sequence should be equal to or greater than the number of dimensions of the reducer.

template<class S>
bool broadcast_shape(S &shape, bool reuse_cache = false) const

Broadcast the shape of the reducer to the specified parameter.

Return

a boolean indicating whether the broadcasting is trivial

Parameters
• shape: the result shape

• reuse_cache: parameter for internal optimization

template<class S>
bool has_linear_assign(const S &strides) const

Checks whether the xreducer can be linearly assigned to an expression with the specified strides.

Return

a boolean indicating whether a linear assign is possible

template<class F, class E, class X, class EVS = std::tuple<evaluation_strategy::lazy_type>, xtl::check_concept<xtl::negation<is_reducer_options<X>>, detail::is_xreducer_functors<F>> = 0>
auto xt::reduce(F &&f, E &&e, X &&axes, EVS &&options = EVS())

Returns an xexpression applying the specified reducing function to an expression over the given axes.

The returned expression either hold a const reference to

e or a copy depending on whether e is an lvalue or an rvalue.
Parameters
• f: the reducing function to apply.

• e: the xexpression to reduce.

• axes: the list of axes.

• options: evaluation strategy to use (lazy (default), or immediate)