xmanipulation

Defined in xtensor/xmanipulation.hpp

template<std::size_t N, class E>
auto xt::atleast_Nd(E &&e)

Expand dimensions of xexpression to at least N

This adds newaxis() slices to a strided_view until the dimension of the view reaches at least N. Note: dimensions are added equally at the beginning and the end. For example, a 1-D array of shape (N,) becomes a view of shape (1, N, 1).

Return

strided_view with expanded dimensions

Parameters

• e: input xexpression

Template Parameters

• N: the number of requested dimensions

template<class E>
auto xt::atleast_1d(E &&e)

Expand to at least 1D.

See

atleast_Nd

template<class E>
auto xt::atleast_2d(E &&e)

Expand to at least 2D.

See

atleast_Nd

template<class E>
auto xt::atleast_3d(E &&e)

Expand to at least 3D.

See

atleast_Nd

template<class E>
auto xt::expand_dims(E &&e, std::size_t axis)

Expand the shape of an xexpression.

Insert a new axis that will appear at the axis position in the expanded array shape. This will return a strided_view with a xt::newaxis() at the indicated axis.

Return

returns a strided_view with expanded dimension

Parameters

• e: input xexpression

• axis: axis to expand

template<layout_type L = xt::layout_type::row_major, class E>
auto xt::flatten(E &&e)

Returns a flatten view of the given expression.

No copy is made. This method is equivalent to ravel and is provided for API sameness with Numpy.

See

ravel

Parameters

• e: the input expression

Template Parameters

• L: the layout used to read the elements of e. If no parameter is specified, XTENSOR_DEFAULT_TRAVERSAL is used.

• E: the type of the expression

template<layout_type L, class T>
auto xt::flatnonzero(const T &arr)

return indices that are non-zero in the flattened version of arr, equivalent to nonzero(ravel<layout_type>(arr))[0];

Return

indices that are non-zero in the flattened version of arr

Parameters

• arr: input array

template<class E>
auto xt::flip(E &&e, std::size_t axis)

Reverse the order of elements in an xexpression along the given axis.

Note: A NumPy/Matlab style flipud(arr) is equivalent to xt::flip(arr, 0), fliplr(arr) to xt::flip(arr, 1).

Return

returns a view with the result of the flip

Parameters

• e: the input xexpression

• axis: the axis along which elements should be reversed

template<layout_type L = xt::layout_type::row_major, class E>
auto xt::ravel(E &&e)

Returns a flatten view of the given expression.

No copy is made.

Parameters

• e: the input expression

Template Parameters

• L: the layout used to read the elements of e. If no parameter is specified, XTENSOR_DEFAULT_TRAVERSAL is used.

• E: the type of the expression

template<class E>
auto xt::repeat(E &&e, std::size_t repeats, std::size_t axis)

Repeats elements of an expression along a given axis.

Return

an expression which as the same shape as e, except along the given axis

Parameters

• e: the input xexpression

• repeats: The number of repetition of each elements. repeats is broadcasted to fit the shape of the given axis.

• axis: the axis along which to repeat the value

template<class E>
auto xt::repeat(E &&e, const std::vector<std::size_t> &repeats, std::size_t axis)

Repeats elements of an expression along a given axis.

Return

an expression which as the same shape as e, except along the given axis

Parameters

• e: the input xexpression

• repeats: The number of repetition of each elements. The size of repeats must match the shape of the given axis.

• axis: the axis along which to repeat the value

template<class E>
auto xt::repeat(E &&e, std::vector<std::size_t> &&repeats, std::size_t axis)

Repeats elements of an expression along a given axis.

Return

an expression which as the same shape as e, except along the given axis

Parameters

• e: the input xexpression

• repeats: The number of repetition of each elements. The size of repeats must match the shape of the given axis.

• axis: the axis along which to repeat the value

template<class E>
auto xt::roll(E &&e, std::ptrdiff_t shift)

Roll an expression.

The expression is flatten before shifting, after which the original shape is restore. Elements that roll beyond the last position are re-introduced at the first. This function does not change the input expression.

Return

a roll of the input expression

Parameters

• e: the input xexpression

• shift: the number of places by which elements are shifted

template<class E>
auto xt::roll(E &&e, std::ptrdiff_t shift, std::ptrdiff_t axis)

Roll an expression along a given axis.

Elements that roll beyond the last position are re-introduced at the first. This function does not change the input expression.

Return

a roll of the input expression

Parameters

• e: the input xexpression

• shift: the number of places by which elements are shifted

• axis: the axis along which elements are shifted.

template<std::ptrdiff_t N = 1, class E>
auto xt::rot90(E &&e, const std::array<std::ptrdiff_t, 2> &axes = {0, 1})

Rotate an array by 90 degrees in the plane specified by axes.

Rotation direction is from the first towards the second axis.

Return

returns a view with the result of the rotation

Parameters

• e: the input xexpression

• axes: the array is rotated in the plane defined by the axes. Axes must be different.

Template Parameters

• N: number of times the array is rotated by 90 degrees. Default is 1.

template<class E>
auto xt::split(E &e, std::size_t n, std::size_t axis = 0)

Split xexpression along axis into subexpressions.

This splits an xexpression along the axis in n equal parts and returns a vector of strided_view. Calling split with axis > dimension of e or a n that does not result in an equal division of the xexpression will throw a runtime_error.

Parameters

• e: input xexpression

• n: number of elements to return

• axis: axis along which to split the expression

template<class E>
auto xt::hsplit(E &e, std::size_t n)

Split an xexpression into subexpressions horizontally (column-wise)

This method is equivalent to split(e, n, 1).

Parameters

• e: input xexpression

• n: number of elements to return

template<class E>
auto xt::vsplit(E &e, std::size_t n)

Split an xexpression into subexpressions vertically (row-wise)

This method is equivalent to split(e, n, 0).

Parameters

• e: input xexpression

• n: number of elements to return

template<class E>
auto xt::squeeze(E &&e)

Returns a squeeze view of the given expression.

No copy is made. Squeezing an expression removes dimensions of extent 1.

Parameters

• e: the input expression

Template Parameters

• E: the type of the expression

template<class E, class S, class Tag = check_policy::none, std::enable_if_t<!xtl::is_integral<S>::value, int> = 0>
auto xt::squeeze(E &&e, S &&axis, Tag check_policy = Tag())

Remove single-dimensional entries from the shape of an xexpression.

Parameters

• e: input xexpression

• axis: integer or container of integers, select a subset of single-dimensional entries of the shape.

• check_policy: select check_policy. With check_policy::full(), selecting an axis which is greater than one will throw a runtime_error.

template<class E>
auto xt::transpose(E &&e)

Returns a transpose view by reversing the dimensions of xexpression e.

Parameters

• e: the input expression

template<class E, class S, class Tag = check_policy::none>
auto xt::transpose(E &&e, S &&permutation, Tag check_policy = Tag())

Returns a transpose view by permuting the xexpression e with permutation.

Parameters

• e: the input expression

• permutation: the sequence containing permutation

• check_policy: the check level (check_policy::full() or check_policy::none())

Template Parameters

• Tag: selects the level of error checking on permutation vector defaults to check_policy::none.

template<class E>
auto xt::trim_zeros(E &&e, const std::string &direction = "fb")

Trim zeros at beginning, end or both of 1D sequence.

Return

returns a view without zeros at the beginning and end

Parameters

• e: input xexpression

• direction: string of either ‘f’ for trim from beginning, ‘b’ for trim from end or ‘fb’ (default) for both.