"User Guide"

User Guide

Practical notes for using the current Stratax API from C++ and Python.

Containers

Stratax centers on Vector, Matrix, and Tensor.

#include <stratax.h>
 
stratax::Vector<double> values{1.0, 2.0, 3.0};
auto doubled = values * 2.0;

from stratax import Vector
 
values = Vector([1.0, 2.0, 3.0])
doubled = values * 2.0

Use Vector for one-dimensional data, Matrix for two-dimensional data, and Tensor for N-dimensional data. Shape stores dimension metadata for reshape and tensor construction.

Indexing and Slicing

C++ uses operator() for multidimensional access and slice(...) helpers for slicing.

Core container types use operator[] for flat indexing making container types share algorithm logic.

stratax::Vector<double> values{1.0, 2.0, 3.0, 4.0};
values(1) = 5.0;
auto every_other = stratax::slice(values, stratax::Slice{0, values.size(), 2});

Python uses familiar item access and slice syntax.

from stratax import Matrix
 
matrix = Matrix([[1.0, 2.0], [3.0, 4.0]])
matrix[1, 0] = 9.0
first_column = matrix[:, 0]
reversed_rows = matrix[::-1, :]

Negative indexing is supported in both APIs. Slicing currently returns copy-based results.

Shape, Reshape, and Flatten

Use Shape when constructing tensors or changing logical dimensions.

stratax::Tensor<double> tensor(stratax::Shape{2, 2, 2}, 1.0);
auto matrix_like = stratax::reshape(tensor, stratax::Shape{4, 2});
auto flat = stratax::flatten(tensor);

from stratax import Shape, Tensor
 
tensor = Tensor(Shape([2, 2, 2]), 1.0)
matrix_like = tensor.reshape(Shape([4, 2]))
flat = tensor.flatten()

Reshape requires the total element count to stay compatible with the source container.

Creation and Conversion

Creation helpers provide common initialized containers.

auto zeros_tensor = stratax::zeros<double>(stratax::Shape{2, 3});
auto eye = stratax::identity<double>(3);

from stratax import Shape, full, identity, ones, zeros
 
zeros_tensor = zeros(Shape([2, 3]))
ones_vector = ones(4)
full_matrix = full(2, 3, 7.0)
eye = identity(3)

Conversions move between compatible container shapes.

auto tensor = stratax::to_tensor(values);
auto roundtrip = stratax::to_vector(tensor);

from stratax import to_tensor, to_vector
 
tensor = to_tensor(values)
roundtrip = to_vector(tensor)

Reductions

Reductions are available globally and by axis.

auto total = stratax::sum(tensor);
auto last_axis = stratax::reductions::sum(tensor, -1);
auto mean_axis0 = stratax::reductions::mean(tensor, 0, true);

from stratax import mean, sum
 
total = sum(tensor)
last_axis = sum(tensor, -1)
mean_axis0 = mean(tensor, 0, keepdims=True)

Python reductions support keepdims for preserving rank.

Python API Notes

Python bindings currently expose double container specializations. Use tolist() for quick inspection or for handing values to ordinary Python code.

from stratax import Vector
 
values = Vector([1.0, 2.0, 3.0])
print(values.tolist())

For a compact list of Python constructors, helpers, reductions, and exceptions, see Python API Reference.

Current Limits

Non-owning views, linear algebra, random sampling, statistics, and file I/O are planned areas. Arithmetic broadcasting is available; comparisons remain shape-sensitive. Check Roadmap for the current direction.