Coordinate Vector

In the worlds of mathematics and physics, a vector represents a fundamental concept—an arrow possessing both magnitude and direction. It exists as an independent entity, whether describing the velocity of a planet, the force on a bridge, or a displacement in space. However, to work with this abstract object, we must describe it with numbers, and this description is not unique. The specific set of numbers we use, known as a ​​coordinate vector​​, depends entirely on the frame of reference, or "basis," we choose. This raises a critical question: how do we reconcile these different numerical descriptions of the same underlying reality, and how can we leverage this flexibility to our advantage?

This article demystifies the coordinate vector, bridging the gap between its abstract definition and its powerful applications. We will explore how changing our mathematical perspective is not just a theoretical exercise but a practical tool for solving complex problems. You will learn how the same physical vector can be represented by different coordinates and how to translate between these descriptions. The article is structured to guide you from foundational concepts to advanced applications. In "Principles and Mechanisms," we will dissect the core ideas of basis, coordinate representation, and the mechanics of transformation. Following this, "Applications and Interdisciplinary Connections" will demonstrate how these principles are applied across physics and engineering, from analyzing material stress to understanding the fabric of spacetime in general relativity.

Imagine you're trying to describe a location to a friend. You might say, "From the town square, walk three blocks East and then two blocks North." This is a perfectly clear set of instructions. The numbers $(3, 2)$ are the ​​coordinates​​, and the directions (East, North) are the ​​basis​​ of your description. But what if your friend is a sailor who navigates by landmarks on the coast? They might prefer instructions like, "Follow the coastline for 5 kilometers, then turn perpendicular to the shore and head inland for 1 kilometer." The destination is the same, but the coordinates and the basis are completely different.

This simple idea is the heart of what a ​​coordinate vector​​ is in mathematics and physics. A vector is a geometric object—an arrow with a specific length and direction, representing a displacement, a force, or a velocity. It exists independently of how we choose to describe it. The coordinate vector is simply the recipe for constructing that vector using a chosen set of basis vectors. It's the list of numbers that tells us "how much" of each basis vector we need to add together to get our final arrow.

Let's make this more concrete. In a familiar two-dimensional plane, we often use the ​​standard basis​​, which consists of two perpendicular vectors of unit length, typically called $\vec{e}_1$ and $\vec{e}_2$. You can think of them as "one step to the right" and "one step up" on a piece of graph paper. A vector $\vec{v} = \begin{pmatrix} 3 \\ 1 \end{pmatrix}$ in this system means simply $\vec{v} = 3\vec{e}_1 + 1\vec{e}_2$.

But we are free to choose any set of non-parallel vectors as our basis. Suppose we define a new basis, $B = \{\vec{b}_1, \vec{b}_2\}$, where our new instructions are given in terms of different directions and step sizes. For instance, let $\vec{b}_1 = \begin{pmatrix} 3 \\ -1 \end{pmatrix}$ and $\vec{b}_2 = \begin{pmatrix} -2 \\ 4 \end{pmatrix}$. Now, if we are given a vector $\vec{u}$ with its coordinates in this new basis, written as $[\vec{u}]_B = \begin{pmatrix} 5 \\ 2 \end{pmatrix}$, what does this mean? It's a recipe: take 5 steps in the $\vec{b}_1$ direction and 2 steps in the $\vec{b}_2$ direction. The true vector $\vec{u}$, in the familiar standard language, is found by simply following these instructions:

So, the coordinate vector $[\vec{u}]_B = \begin{pmatrix} 5 \\ 2 \end{pmatrix}$ represents the same physical vector as $\begin{pmatrix} 11 \\ 3 \end{pmatrix}$ in the standard system. The vector itself is unchanged; only our description of it has been altered.

The wonderful thing about a coordinate system is that once you've chosen your basis, the rules of algebra work beautifully. If you have a vector $\vec{v}$ with coordinates $[\vec{v}]_B = \begin{pmatrix} \alpha \\ \beta \end{pmatrix}$, and you want to find the coordinates of a new vector $\vec{w} = \vec{v} - \vec{b}_1$, you don't need to convert back to the standard system. The recipe for $\vec{v}$ is $\alpha\vec{b}_1 + \beta\vec{b}_2$. So, the recipe for $\vec{w}$ is simply $(\alpha\vec{b}_1 + \beta\vec{b}_2) - \vec{b}_1 = (\alpha - 1)\vec{b}_1 + \beta\vec{b}_2$. Its coordinates in the basis $B$ are just $[\vec{w}]_B = \begin{pmatrix} \alpha - 1 \\ \beta \end{pmatrix}$. The algebra of vectors becomes the simple algebra of their components.

This process of converting from a special basis to the standard one is a common task. Imagine a robotics engineer calibrating a drone. The lab has a fixed "world" coordinate system (North, East, Up), but the drone has its own internal system (Forward, Right, Up) that rotates as it flies. A sensor on the drone reports an object's position as $[v]_B = \begin{pmatrix} 5 \\ 2 \\ -1 \end{pmatrix}$ in its own basis $B$. To know where that object is in the lab, the engineer must translate.

This translation has a wonderfully elegant structure. If we have a basis $B = \{\vec{b}_1, \vec{b}_2, \dots, \vec{b}_n\}$ and a coordinate vector $[\vec{v}]_B = (c_1, c_2, \dots, c_n)^T$, the vector $\vec{v}$ in the standard basis is:

This is a linear combination, which can be expressed as a matrix multiplication. If we build a ​​change-of-basis matrix​​, $P$, whose columns are simply the basis vectors of $B$ written in standard coordinates, then the translation is just $\vec{v} = P [\vec{v}]_B$.

This matrix $P$ acts as a universal translator or a dictionary. For a 2D video game rover whose "forward" is $\vec{b}_1$ and "right" is $\vec{b}_2$, the matrix $P = \begin{pmatrix} \vec{b}_1 \vec{b}_2 \end{pmatrix}$ instantly converts any local movement command, like "move 5 units forward and 2 units right," into the corresponding movement in the game's world coordinates.

What about the other way around? Suppose you want to perform a standard action, like moving purely horizontally in a game world, but your game engine uses a skewed, non-orthogonal basis for some artistic effect. You have the vector in the standard basis, $\vec{v} = \begin{pmatrix} 1 \\ 0 \end{pmatrix}$, and you need to find its coordinates $[c_1, c_2]^T$ in the new basis $B' = \{\vec{b}_1, \vec{b}_2\}$.

This is the inverse problem. We are looking for the unknown coefficients in the equation $\vec{v} = c_1 \vec{b}_1 + c_2 \vec{b}_2$. This is a system of linear equations. Using our change-of-basis matrix $P$ from before, the equation is $\vec{v} = P [\vec{v}]_{B'}$. To find the unknown coordinates $[\vec{v}]_{B'}$, we need to "undo" the action of $P$. This is accomplished by the inverse matrix, $P^{-1}$.

The existence of this inverse translator is guaranteed as long as our basis vectors are not collinear—that is, as long as they form a genuine basis capable of spanning the whole space.

With all these components changing from one perspective to another, you might wonder if anything is absolute. Yes, there is: the zero vector, $\vec{0}$. No matter what basis you choose—orthogonal, skewed, or wildly imaginative—the coordinates of the zero vector are always $(0, 0, \dots, 0)$. Why?

The reason goes to the very definition of a basis. A set of vectors $\{\vec{b}_1, \dots, \vec{b}_n\}$ forms a basis only if they are ​​linearly independent​​. This is a powerful statement. It means that the only way to combine these vectors to get the zero vector is by taking zero amount of each:

Therefore, the unique recipe for constructing the zero vector is to use zero parts of every basis vector. This makes the origin a universal landmark, an anchor point agreed upon by all observers, regardless of their coordinate system.

So far, our basis vectors have been constant; the "East" direction is the same everywhere. But what if our coordinate grid itself is curved, like the lines of longitude and latitude on a globe, or the concentric circles and radial lines of a polar coordinate system?

Here, we enter the world of differential geometry. A vector is no longer just an arrow you can slide around; it becomes a ​​tangent vector​​, an arrow attached to a specific point on a surface or in a space. Its components now depend not just on the coordinate system, but also on the location. In polar coordinates $(r, \theta)$, the "r-direction" (outward from the origin) and "$\theta$-direction" (counter-clockwise) are different at every single point.

How do the components of a vector transform in this case? The change-of-basis matrix is replaced by the ​​Jacobian matrix​​, which involves partial derivatives. For a vector with Cartesian components $(V^x, V^y)$, its components $(V^r, V^\theta)$ in the polar system are found by a rule that essentially asks: "At this point, how much does a small step in the x-direction contribute to movement in the r-direction and $\theta$-direction?". The transformation law becomes local, a dynamic translation that changes from point to point.

This is where the magic happens. If a vector's components are so fluid, changing with every choice of coordinates, what is the "real" vector? The answer is that the physical properties of the vector, like its length (or magnitude), must be ​​invariant​​. They cannot depend on the language we use to describe them.

Let's take a vector at the point $(x,y)=(4,3)$ with Cartesian components $V^x=2$ and $V^y=-3$. Its squared magnitude is trivial to calculate: $\|\vec{V}\|^2 = (V^x)^2 + (V^y)^2 = 2^2 + (-3)^2 = 13$.

Now, let's do this the hard way. We transform to polar coordinates. At $(4,3)$, we find $r=5$. After applying the Jacobian transformation, we find the polar components are something completely different, like $V^r = -1/5$ and $V^\theta = -18/25$. These numbers look nothing like our original components.

But if we try to calculate the magnitude in polar coordinates, we must use the "ruler" appropriate for that system. The formula for squared length in polar coordinates is not just the sum of squares; it's $(V^r)^2 + r^2(V^\theta)^2$. The $r^2$ factor is part of the ​​metric tensor​​, $g_{ij}$, which defines geometry in a given coordinate system. Plugging in our values:

The result is exactly the same. This is a profound revelation. The components are shadows, the metric is the ruler, but the length is the reality. Physics is built on such invariants—quantities that all observers, no matter their coordinate system, can agree upon.

To cap off our journey, there's one final, beautiful twist. It turns out that there are two complementary ways for vector components to transform. The vectors we've discussed so far, like displacement or velocity, are called ​​contravariant vectors​​. Their components transform using the Jacobian matrix, let's call it $A$.

But there's another type of object, a ​​covariant vector​​ (or covector), which is often used to represent things like forces or gradients. Its components transform in a different, "dual" way. If the contravariant components transform via a matrix $A$, the covariant components transform via the matrix $B = (A^{-1})^T$, the transpose of the inverse of $A$.

This isn't just mathematical pedantry. This dual transformation rule is precisely what's needed to ensure that when you combine a covariant vector (like force, $\vec{F}$) with a contravariant vector (like displacement, $d\vec{x}$), the result—the scalar product representing work done, $\vec{F} \cdot d\vec{x}$—is an invariant scalar. It's another layer of the universe's internal consistency, ensuring that physical realities don't depend on the mathematical language we invent to describe them. The coordinate vector, which began as a simple recipe, has led us to the deep and unified structure of geometry itself.

After our exploration of the principles behind coordinate vectors, you might be left with a nagging question: "This is all fine and good as a mathematical game, but what is it for?" It's a fair question, and the answer is wonderfully far-reaching. The idea that a single, unchanging physical reality—a force, a velocity, a field—can be represented by different sets of numbers depending on our point of view is one of the most powerful tools in the physicist's and engineer's toolkit. It’s like understanding that the shadow an object casts depends on where you shine the light from, but the object itself remains the same. Learning to work with coordinate vectors is learning the art of choosing the best light source for the job.

In this chapter, we’ll take a journey through some of these applications, seeing how this one simple concept provides the language to describe everything from the deformation of steel beams to the strange world of an accelerating astronaut and the very fabric of spacetime.

Let's start on familiar ground. Imagine you're in a laboratory and you measure a force vector. You write down its components—say, how much force is directed along the north-south line, the east-west line, and the up-down line. Now, what happens if your colleague in the same lab has set up their measurement apparatus rotated by some angle? They will measure the same force, the same physical arrow hanging in space, but they will write down a completely different set of numbers.

The rules that translate your numbers into your colleague's numbers are the rules of coordinate transformation. For a simple rotation, the new components are a specific "mix" of the old ones, prescribed by the mathematics of sines and cosines. While the individual components change, the underlying physical object does not. Its length, for instance, remains stubbornly the same, a fact guaranteed by the geometric properties of the rotation.

But not all transformations are so gentle. In engineering and materials science, we are often interested in how things bend, stretch, and deform. Imagine a block of rubber. If you push on the top surface parallel to the bottom, the block deforms in what's called a ​​shear​​. This is not a rigid rotation; the shape of the block itself changes. A vector embedded in this rubber would be stretched and tilted. A shear transformation describes how the components of this vector would change, and unlike a rotation, it alters the vector's length and its angle relative to other vectors. Understanding these transformations allows engineers to precisely model the stress and strain on materials, which is rather important if you want the bridges you design to stay standing.

Physics and engineering are also a creative endeavor; we constantly build new physical quantities out of ones we already know. For instance, we might define a new vector, let's call it $\vec{V}$, by combining a known vector $\vec{u}$ with a more complex object called a tensor $\vec{T}$, perhaps representing the stress in a material or the electromagnetic field. A key question is, does our new creation, defined by $V_j = u_i T_{ij}$, still behave like a proper vector? That is, if we rotate our coordinate system, will its new components $V'_j$ be correctly predicted by the standard vector transformation rules? The mathematics of tensors provides a definitive "yes." By following the transformation laws for $\vec{u}$ and $\vec{T}$, we can prove that the resulting object $\vec{V}$ transforms exactly as a vector should, ensuring our new quantity is a physically meaningful concept and not just a meaningless jumble of numbers.

The rectangular grid of a Cartesian coordinate system is a comfortable place, but nature is rarely so accommodating. Many problems in physics have a natural symmetry—the circular motion of a planet, the spherical spread of a sound wave, the cylindrical flow of water in a pipe—that makes Cartesian coordinates awkward and clumsy. The obvious solution is to adopt a coordinate system that respects the symmetry of the problem, like polar, cylindrical, or spherical coordinates.

This choice, however, introduces a fascinating and profound new feature. Think about the basis vectors in a polar coordinate system $(r, \theta)$. The basis vector $\hat{e}_r$ always points radially away from the origin, and $\hat{e}_\theta$ points in the direction of increasing angle. Now, pick two different points in the plane. The $\hat{e}_r$ at the first point points in a different direction than the $\hat{e}_r$ at the second! Our very "rulers" for measuring vector components change from place to place.

Let's see what this means with a concrete example. Imagine a wide, steady river flowing uniformly due east. In a Cartesian system with the x-axis pointing east, the velocity vector is the same everywhere: a simple constant, $\vec{V} = (V_0, 0, 0)$. Now, let's try to describe this same, simple flow using cylindrical coordinates. Suddenly, the components of the velocity, $V^\rho$ and $V^\phi$, are no longer constant. They become functions of the angle $\phi$. This isn't because the water has suddenly started to swirl; the flow is still perfectly uniform. The complexity has arisen entirely from our choice of description—from the fact that our coordinate basis vectors are themselves rotating as we move around the origin.

This leads to a crucial question: how do we talk about the rate of change of a vector field in these curvilinear systems? If we just take the ordinary partial derivative of the components, we'll be misled. We'll calculate a non-zero change even for a constant field, simply because our basis vectors are changing. Physics requires a smarter tool: the ​​covariant derivative​​. The covariant derivative, $\nabla_j A^i$, starts with the simple partial derivative of the components, $\frac{\partial A^i}{\partial x^j}$, but adds a special correction term involving objects called Christoffel symbols. This correction term is not just some mathematical formalism; it has a deep physical job. It precisely accounts for the way the basis vectors are twisting and turning as we move from point to point. For that uniform force field, the covariant derivative correctly tells us that the physical field is not changing—the Christoffel symbol term exactly cancels the change we see in the components, revealing the unchanging physical reality underneath our choice of coordinates.

The journey doesn't end there. As we venture into the realms of modern physics, like Einstein's theory of general relativity, our concept of a vector and its coordinates deepens even further into powerful abstractions.

One such abstraction is to think of a vector not as an "arrow," but as an ​​operator​​—a machine that performs a specific task. What task? It measures the rate of change of any scalar quantity (like temperature or pressure) in a particular direction. From this perspective, a vector $V$ acting on a scalar field $f$ produces a new scalar, $V(f)$, the directional derivative. The vector's components, $V^i$, are then seen in a new light: they are simply the instructions telling the operator how much to "differentiate along" each coordinate direction. This viewpoint is the bedrock of differential geometry, the mathematical language of general relativity.

In the strange, curved spacetimes of general relativity, coordinate systems can be particularly unwieldy. The basis vectors may be non-orthogonal and have varying lengths, making even a simple dot product a chore. Physicists, being practical people, developed a clever workaround. At any point in spacetime, they define a small, private, perfectly flat and orthonormal reference frame—a set of ideal rulers. In this local frame, the laws of special relativity apply, and calculations are simple. The components of a vector in this "nice" local frame are then related to the components in the "messy" global coordinate system by a translator matrix called a ​​vielbein​​ or frame field. The vielbein acts as a dictionary, allowing one to switch back and forth between the convenient physical picture (the local orthonormal frame) and the necessary mathematical description (the global coordinates).

This ability to translate between viewpoints is paramount when comparing the experiences of different observers. Consider an astronaut in a rocket accelerating uniformly through otherwise empty space. Their description of spacetime, given by ​​Rindler coordinates​​, is fundamentally different from that of an inertial observer floating freely. A fundamental symmetry for the inertial observer—the fact that the laws of physics are the same today as they are tomorrow (time-translation invariance)—is represented by a simple constant vector. But when we translate the components of this vector into the Rindler coordinates of the accelerating astronaut, they become complicated functions of position and time. This transformation reveals a profound physical insight: concepts like energy, which is conserved due to this time symmetry for an inertial observer, become ambiguous and non-conserved from the perspective of an accelerating observer.

Finally, it's worth noting that the world of vectors is richer than we've let on. For every type of vector we've discussed (technically called a contravariant vector), there exists a dual object called a covector (or covariant vector). These objects also have components in a basis, but they transform according to a slightly different rule. This distinction, while subtle in simple flat spaces, becomes absolutely critical in the curved geometries of general relativity, where vectors and covectors represent genuinely different kinds of physical objects.

From the simple rotation of a measurement device to the warping of spacetime near a black hole, the concept of a coordinate vector is the common thread. We have seen how it provides the essential language for describing change, whether it's the physical deformation of a solid, the mathematical description of a uniform flow in a new geometry, or the apparent change in physical laws for an accelerating observer.

The central lesson is one of the most beautiful in all of physics: the importance of distinguishing the essential, invariant physical reality from the arbitrary, observer-dependent description we give it. The numbers we write down—the components—are our choice. The underlying physics is not. The art of theoretical physics is largely the art of skillfully translating between these different descriptions to find the one in which the physical reality is revealed in its simplest and most elegant form.