Fiber of a map - Ciallo Wiki

1 Definition and basic properties

1.1 Preimage and fiber over a point

Let \(f: X \to Y\) be a function between sets. For a point \(y \in Y\), the fiber over \(y\) is the preimage of \(y\): \[ f^{-1}(y)=\{x \in X : f(x)=y\}\subseteq X. \] Fibers collect all domain points that map to the same target value. Conceptually, they “slice” the domain according to the value taken by \(f\).

1.2 Level sets as fibers

For a fixed \(y\in Y\), the fiber \(f^{-1}(y)\) is also a level set: it consists of points where the function attains the prescribed value. When \(Y\) is a subset of \(\mathbb{R}\), level sets are often interpreted geometrically as curves, surfaces, or discrete sets.

1.3 Fibers for set-valued comparisons (empty, singleton, and general cases)

Fibers can vary in size:

Empty fibers: If no \(x\in X\) satisfies \(f(x)=y\), then \(f^{-1}(y)=\varnothing\).
Singleton fibers: If exactly one \(x\) maps to \(y\), then \(f^{-1}(y)=\{x\}\).
General fibers: If multiple points map to the same \(y\), the fiber is a set with multiple elements.

These cases reflect how strongly (or weakly) \(f\) distinguishes points of \(X\).

1.4 Fibers under composition of maps

Given composable maps \(X \xrightarrow{f} Y \xrightarrow{g} Z\), consider \(h=g\circ f: X \to Z\). For \(z\in Z\), \[ h^{-1}(z)=\{x\in X : g(f(x))=z\}. \] This can be expressed in terms of fibers of \(g\) as a union of fibers of \(f\): \[ (g\circ f)^{-1}(z)=\bigcup_{y\in g^{-1}(z)} f^{-1}(y). \] Thus, the fiber of a composite map is assembled from fibers of the intermediate map(s).

2 Fibers and map properties

2.1 Injectivity and singleton fibers

A function \(f: X\to Y\) is injective if different points of \(X\) never share the same image.

2.1.1 Characterization via fibers of points in the image

A standard characterization is:

\(f\) is injective iff every fiber over a point in the image is either a singleton or empty; equivalently, for all \(y\in f(X)\), the fiber \(f^{-1}(y)\) has exactly one element.

If injectivity holds, each \(y\) that is attained corresponds to exactly one \(x\).

2.2 Surjectivity and nonempty fibers

A function is surjective if every target point is hit by at least one domain point.

2.2.1 Behavior of fibers outside the image

If \(f\) is not surjective, then some fibers are empty. In general:

For \(y\in f(X)\), the fiber \(f^{-1}(y)\) is nonempty.
For \(y\notin f(X)\), the fiber is empty.

So surjectivity is equivalent to all fibers being nonempty.

2.3 Constant maps and fibers equal to the whole domain

If \(f:X\to Y\) is constant, meaning \(f(x)=y_0\) for all \(x\in X\), then:

\(f^{-1}(y_0)=X\),
\(f^{-1}(y)=\varnothing\) for \(y\neq y_0\).

In this case, the map collapses the entire domain into a single target value, and there is only one nontrivial fiber.

2.4 Fibers and equivalence relations induced by maps

A function can induce a relation on the domain. Define \(x\sim x'\) iff \(f(x)=f(x')\). This is an equivalence relation, and its equivalence classes are precisely the fibers: \[ [x]=f^{-1}(f(x)). \] Thus, fibers describe the partition of \(X\) into groups of points that the map identifies.

3 Topological aspects (in analysis context)

3.1 Fibers and continuity via inverse images

Suppose \(X\) and \(Y\) are topological spaces. If \(f:X\to Y\) is continuous, then for any open set \(U\subseteq Y\), the preimage \(f^{-1}(U)\) is open in \(X\). Fibers are particular preimages: \[ f^{-1}(y)=f^{-1}(\{y\}). \] Whether a fiber is open or closed depends on the topological nature of the singleton \(\{y\}\) in \(Y\).

3.2 Closed/open fibers under additional assumptions

If the singleton \(\{y\}\) is closed in \(Y\) (for instance, when \(Y\) is a Hausdorff space), then continuity implies:

\(f^{-1}(y)\) is closed in \(X\).

If \(\{y\}\) is open (a rarer situation, depending on the topology), then the fiber is open as well.

3.2.1 Relationship with the topology on the target space

This dependence highlights a general principle: topological properties of fibers reflect how the target space treats individual points. For example, in a discrete topology on \(Y\), every singleton is open and closed, so every fiber becomes clopen in \(X\) when \(f\) is continuous.

3.3 Fibers of restrictions and subspaces

When \(A\subseteq X\) and one considers the restriction \(f	_A: A\to Y\), fibers change by intersection:

(f	_A)^{-1}(y)=A\cap f^{-1}(y).

\] So the fiber inside a subspace is the portion of the original fiber lying in that subspace.

3.4 Fiber structure under homeomorphisms and embeddings

If \(f:X\to Y\) and one applies a homeomorphism \(u:Y\to Y'\), the composite map \(u\circ f:X\to Y'\) has fibers related by relabeling. Specifically, if \(y'=u(y)\), then \[ (u\circ f)^{-1}(y')=f^{-1}(y). \] For embeddings, topological features like closedness, compactness, and connectedness of fibers can often be transferred between subspaces and their ambient spaces, depending on how the embedding interacts with the relevant subsets.

4 Measure and regularity viewpoints

4.1 Fibers and measurability of inverse images

In measurable settings, one often studies whether fibers are measurable sets. If \(f\) is measurable and \(\{y\}\) is measurable in the \(\sigma\)-algebra of \(Y\), then \(f^{-1}(y)\) is measurable in \(X\). In many standard contexts (e.g., Borel \(\sigma\)-algebras on \(\mathbb{R}^n\)), singletons are measurable, making fibers measurable for Borel-measurable maps.

4.2 Fiber size and “typical” behavior (heuristic overview)

A recurring theme in analysis is that fibers can behave very differently depending on the target value \(y\). In many regular problems, “most” values exhibit fibers of controlled size (for example, finite or manifold-like), while exceptional values correspond to singular behavior. This motivates questions about how fiber cardinality, measure, or dimension varies across \(Y\).

4.3 Regular values and smooth maps (intuition-focused)

Consider a smooth map between manifolds. Informally, a regular value is a point \(y\) such that the map behaves like a submersion near every \(x\) mapping to \(y\). Under that assumption, the preimage \(f^{-1}(y)\) typically forms a smooth submanifold of the appropriate dimension.

4.3.1 Dimension/counting intuition for fibers

A heuristic for smooth maps \(f:\mathbb{R}^n\to\mathbb{R}^m\) is:

If the map is locally “full rank” at points of the fiber, then the constraints \(f(x)=y\) impose \(m\) independent conditions.
As a result, the fiber often has dimension \(n-m\) near those points.

This intuition helps explain why level sets frequently look like curves/surfaces in geometric analysis.

5 Examples and computation

5.1 Linear maps between Euclidean spaces

Let \(f:\mathbb{R}^n\to\mathbb{R}^m\) be linear, such as \(f(x)=Ax\) for a matrix \(A\). For a given \(y\in\mathbb{R}^m\), the fiber is the solution set of the linear system \[ Ax=y. \]

5.1.1 Fibers as affine subspaces

If \(Ax=y\) is consistent, choose one solution \(x_0\). Then every solution has the form \[ x=x_0+v,\quad v\in \ker(A). \] Hence, a nonempty fiber is an affine translate of \(\ker(A)\), while an inconsistent system yields an empty fiber. This provides an explicit geometric model: fibers are parallel affine spaces.

5.2 Polynomial maps in one variable (level sets)

For a polynomial \(p:\mathbb{R}\to\mathbb{R}\), the fiber over \(y\) is the set of real roots of \(p(x)=y\). Computing it reduces to solving the equation \(p(x)-y=0\). In this setting, fiber cardinality is governed by algebraic facts such as degree and multiplicity of roots.

5.3 Multivariable maps and implicit equation viewpoint

For a function \(f:\mathbb{R}^n\to\mathbb{R}\), fibers are level sets \(f(x)=y\). They can be studied using gradients and local linearization: near points where \(\nabla f\neq 0\), the implicit function theorem viewpoint suggests that the set \(f^{-1}(y)\) behaves like a hypersurface.

5.3.1 Fibers as solution sets of \(f(x)=y\)

More generally, for \(f:\mathbb{R}^n\to\mathbb{R}^m\), the fiber \(f^{-1}(y)\) is the common solution set to the \(m\) equations given by the components of \(f(x)=y\). This interpretation connects fiber computation to solving systems and analyzing the geometry of constraint sets.

6.1 Fibers versus kernels (when the map is linear)

When \(f\) is linear, the kernel \(\ker(f)\) is the fiber over \(0\): \[ \ker(f)=f^{-1}(0). \] Other fibers are affine translates of the kernel whenever they are nonempty. This relationship clarifies why kernels are frequently emphasized in linear algebra as a special fiber case.

6.2 Fibers of quotient maps and induced identification

A quotient map identifies points according to an equivalence relation. In that framework, the fibers correspond to equivalence classes: points in the same class map to the same element of the quotient. Therefore, understanding fibers is central to understanding how quotient constructions partition the original space.

6.3 Fibers in pullback-like settings (conceptual connection)

In categorical or geometric settings, one often forms new spaces by “pulling back” a map along another. Although the detailed formalism can vary, the guiding idea remains: fibers track how points in one space map into another, and pullback constructions refine that tracking by considering compatibility conditions. In many contexts, the resulting preimages can be understood in terms of fibers and how they intersect with the imposed constraints.

7 Applications and connections within analysis

7.1 Solving equations by analyzing fibers

Equations of the form \(f(x)=y\) can be reframed as fiber membership: the solutions are exactly the points in \(f^{-1}(y)\). This approach shifts the focus from solving directly to understanding the structure of the level set.

7.2 Fiber-based decompositions of the domain

Because the fibers partition the domain according to target values, one can decompose \(X\) into disjoint pieces indexed by \(Y\): \[ X=\bigcup_{y\in Y} f^{-1}(y), \quad\text{with disjoint union over attained values.} \] Such decompositions are useful in arguments where one studies the map by examining what happens on each slice.

7.3 Role in studying mapping behavior near points (local slices)

In local analysis, fibers provide “slices” that reveal how \(f\) behaves in a neighborhood. For example, when one investigates stability, regularity, or approximation properties near a point \(x\), the nearby fibers through \(f(x)\) describe how the constraint surface moves and how solutions respond to perturbations.

7.4 Fiber structure in dynamical and iterative settings (broad intuition)

In dynamical systems and iterative processes, one studies how states evolve and how preimages propagate backward through time steps. The set of states that map to a given future state under an iteration is a fiber of the iterate map. Broadly, fiber structure informs questions about how solutions branch, how attractors are organized, and how sensitivity manifests through the geometry of preimage sets.