tensor

History

The idea of tensor calculus originated in the late 19th century. Italian mathematicians Gregorio Ricci-Curbastro and Tullio Levi-Civita created the formal apparatus that later became the language of general relativity. James Clerk Maxwell, without knowing it, had already used tensors in his equations of electromagnetism, and Hendrik Lorentz applied them for transformations of space and time.

How it works

Tensors manifest themselves when we rotate the coordinate system. Ordinary numbers (scalars) do not change upon rotation, while vector components change according to a simple rule. A tensor, however, transforms more complexly, preserving internal geometry, much like the shadow of a rotating cube changes shape but always remains a projection of the same object.

History

The idea of tensor calculus originated in the late 19th century. Italian mathematicians Gregorio Ricci-Curbastro and Tullio Levi-Civita created the formal apparatus that later became the language of general relativity. James Clerk Maxwell, without knowing it, had already used tensors in his equations of electromagnetism, and Hendrik Lorentz applied them for transformations of space and time.

How it works

Tensors manifest themselves when we rotate the coordinate system. Ordinary numbers (scalars) do not change upon rotation, while vector components change according to a simple rule. A tensor, however, transforms more complexly, preserving internal geometry, much like the shadow of a rotating cube changes shape but always remains a projection of the same object.

History

The history of tensors begins with works on differential geometry and mechanics. Gauss and Riemann laid the foundations, but the systematic calculus was developed by Ricci-Curbastro and Levi-Civita in 1884–1900. In 1915, Albert Einstein used their apparatus to formulate general relativity, where gravity manifests as the curvature of spacetime, described by the Ricci tensor and the Einstein tensor. James Clerk Maxwell had already in 1865 essentially written his equations in tensor form, though he didn't use the term. Hendrik Lorentz in 1904 derived transformations that became the basis for understanding the tensor nature of physical laws in special relativity.

How it works

The transformation law for tensor components when transitioning to new coordinates x'^μ = x'^μ(x^ν) is given by the product of Jacobian matrices. For example, for a tensor of rank (1,1): T'^μ_ν = (∂x'^μ/∂x^α) (∂x^β/∂x'^ν) T^α_β. Maxwell's equations in tensor form: ∂_ν F^{μν} = μ_0 J^μ, where F^{μν} is the electromagnetic field tensor, J^μ is the 4-current. Einstein field equations: G_{μν} + Λ g_{μν} = (8πG/c^4) T_{μν}, where G_{μν} is the Einstein tensor expressing spacetime curvature through the metric tensor and its derivatives. Tensors also describe stress and strain in mechanics: σ_{ij} = C_{ijkl} ε^{kl}, where σ is the stress tensor, ε is the strain tensor, and C is the fourth-rank elasticity tensor.

Open problems & fun facts

Developing efficient numerical methods for tensor decompositions of multidimensional data under memory constraints; Finding an exact solution to Einstein's equations for two bodies including gravitational radiation remains a challenge; Mathematical classification of singularities of tensor fields in general relativity

💡 In machine learning, tensors are the basic data type. Neural networks process huge multidimensional tensors: for instance, a color image is a three-dimensional tensor (width, height, color channels), and video is four-dimensional, adding time.