7.13.1 Tensor Component Basis Selection
Tensor Component Basis Selection involves choosing a basis for tensor components to simplify calculations and reveal structural properties in algebraic contexts.
Tensor Component Basis Selection is the preliminary act of choosing a specific basis for the underlying vector space and its dual space, a necessary first step that must precede any extraction of numerical components from an abstractly defined tensor.
Why Selection Must Come First
Tensors Exist Independently of Coordinates
A tensor is defined as a multilinear map without reference to any particular basis, so it possesses no numerical components at all until a basis is chosen; component extraction is not possible without first fixing the frame of reference relative to which those components will be measured.
The Dual Basis Follows Automatically
Once a basis ({e_1, \ldots, e_n}) is selected for (V), the corresponding dual basis ({e^1, \ldots, e^n}) for (V^{*}) is uniquely determined by the biorthogonality condition, so selecting a single basis for (V) is sufficient to fix the frame needed for both vector and covector components.
Criteria Guiding a Good Selection
Adaptation to the Problem's Symmetry
A basis is often chosen to align with any natural symmetry present in the situation being described, such as selecting basis directions along the principal axes of a physical system, since components extracted relative to such an adapted basis frequently take on a simpler, more revealing form than components extracted relative to an arbitrary basis.
Orthonormality for Convenience
When the vector space carries an inner product, selecting an orthonormal basis is a common convenience, since it causes the dual basis to coincide numerically with the original basis under the identification furnished by the inner product, simplifying the relationship between covariant and contravariant components.
Consequences of the Selection for Extraction
Determining Every Subsequent Component Value
Once the basis is selected, the extraction of any component becomes a well-defined operation: evaluating the tensor on the appropriate combination of chosen basis elements. Every numerical entry that will ever appear in the component table is determined jointly by the tensor and this initial selection.
Reversibility of the Selection
Basis selection is never a permanent commitment; a different basis can always be selected afterward, and the transformation law connects the components extracted relative to the two choices, so the initial selection constrains only the immediate numerical form of the components, not the underlying tensor itself.
Selection in Practice
Standard Basis as a Default
In the absence of any special structure or symmetry to exploit, the standard basis, consisting of the elementary unit vectors, is the most common default selection, since it produces the simplest possible dual basis and the most straightforward extraction formulas.
Basis Selection Driven by Application
In applied contexts such as physics or engineering, the basis is frequently selected to match a natural coordinate system already in use for the problem, such as a basis aligned with a rotating frame or with a curved coordinate grid, so that the extracted components carry direct physical interpretation within that setting.
Diagrammatic Illustration
Two different basis selections for the same plane, each producing its own dual basis and therefore its own component readings for the same fixed tensor.
Relationship to the Rest of the Extraction Process
First Step in a Larger Procedure
Basis selection is the opening step in the general procedure of component extraction, followed by evaluating the tensor on the selected basis elements and organizing the resulting values into a component table; without this first step, none of the subsequent steps have any well-defined meaning.
Interplay with Transformation Laws
The specific basis selected also determines the reference point from which any future change-of-basis transformation is measured, so a clearly recorded basis selection is essential not only for the initial extraction of components but also for correctly relating those components to any component set obtained under a different, later selection.