4.3 Natural transformations and functor categories

Natural transformations are the secret sauce of functors, letting us compare them smoothly. They're like bridges between functors, with components that play nice with morphisms. Think of them as the glue that holds the functor world together.

Functor categories take things up a notch, treating functors as objects and natural transformations as morphisms. It's like a meta-category where functors themselves become the stars of the show. This setup lets us study functors in a whole new light.

Natural Transformations

Definition and Components

• Natural transformation $\eta: F \to G$ consists of a family of morphisms $\eta_X: F(X) \to G(X)$ for each object $X$ in the source category
• Components of a natural transformation are the individual morphisms $\eta_X$ that make up the natural transformation
• Components must satisfy the naturality condition for every morphism $f: X \to Y$ in the source category: $G(f) \circ \eta_X = \eta_Y \circ F(f)$
• Naturality condition ensures that the components of the natural transformation are compatible with the morphisms in the source and target categories
• Commutative diagram can be used to visualize the naturality condition

Natural Isomorphisms and Properties

• Natural isomorphism is a natural transformation $\eta: F \to G$ where each component $\eta_X$ is an isomorphism in the target category
• Inverse of a natural isomorphism is also a natural transformation, denoted as $\eta^{-1}: G \to F$
• Identity natural transformation $1_F: F \to F$ has components $(1_F)_X = 1_{F(X)}$ for each object $X$ in the source category
• Composition of natural transformations $\eta: F \to G$ and $\mu: G \to H$ is defined componentwise: $(\mu \circ \eta)_X = \mu_X \circ \eta_X$
• Natural transformations can be used to define equivalence of categories (if there exist natural isomorphisms between two functors)

Functor Categories

Definition and Objects

• Functor category $\mathbf{C^D}$ has functors $F: \mathbf{D} \to \mathbf{C}$ as objects
• Morphisms in the functor category are natural transformations between the functors
• Identity functor $1_\mathbf{D}: \mathbf{D} \to \mathbf{D}$ serves as the identity object in the functor category
• Constant functor $\Delta_X: \mathbf{D} \to \mathbf{C}$ sends every object in $\mathbf{D}$ to a fixed object $X$ in $\mathbf{C}$ and every morphism in $\mathbf{D}$ to the identity morphism $1_X$

Composition of Natural Transformations

• Vertical composition of natural transformations $\eta: F \to G$ and $\mu: G \to H$ is defined as $\mu \circ \eta: F \to H$
• Horizontal composition of natural transformations $\eta: F \to G$ in $\mathbf{C^D}$ and $\theta: H \to K$ in $\mathbf{D^E}$ is a natural transformation $\eta * \theta: F \circ H \to G \circ K$ in $\mathbf{C^E}$
• Horizontal composition is defined componentwise: $(\eta * \theta)_X = \eta_{K(X)} \circ F(\theta_X)$
• Interchange law holds for vertical and horizontal composition: $(\mu \circ \eta) * (\theta \circ \phi) = (\mu * \theta) \circ (\eta * \phi)$

Yoneda Lemma

Statement and Consequences

• Yoneda lemma states that for any functor $F: \mathbf{C^{op}} \to \mathbf{Set}$ and any object $X$ in $\mathbf{C}$, there is a bijection between the set of natural transformations $\text{Nat}(\text{Hom}_\mathbf{C}(X,-), F)$ and the set $F(X)$
• Bijection is given by evaluating a natural transformation at the identity morphism of $X$
• Yoneda lemma implies that the functor $\text{Hom}_\mathbf{C}(X,-)$ (represented functor) completely determines the object $X$ up to isomorphism
• Yoneda embedding $\mathbf{C} \to \mathbf{Set^{C^{op}}}$ sends each object $X$ to its represented functor $\text{Hom}_\mathbf{C}(X,-)$ and each morphism $f: X \to Y$ to the natural transformation $\text{Hom}_\mathbf{C}(f,-)$
• Yoneda embedding is fully faithful, meaning that it preserves and reflects isomorphisms between objects and morphisms