Determinant Geometry: Area, Volume & Orientation

Signed Area in Two Dimensions

The columns of a

2 \times 2

matrix

A

define two vectors in

\mathbb{R}^2

emanating from the origin. These vectors span a parallelogram, and the determinant of

A

equals the signed area of that parallelogram.

For

A = \begin{pmatrix} a & b \\ c & d \end{pmatrix}

, the columns are

\mathbf{u} = (a, c)

and

\mathbf{v} = (b, d)

. The unsigned area of the parallelogram they span can be computed from the cross product magnitude in three dimensions by embedding both vectors with a zero third component, but in two dimensions the formula reduces directly to

Determinant Signed Area 2D

\text{signed area}(\mathbf{u}, \mathbf{v}) = \det\!\begin{pmatrix} u_1 & v_1 \\ u_2 & v_2 \end{pmatrix}

Determinant signed area 2d →

For the matrix above, this evaluates to

ad - bc = \det(A)

.

The sign encodes orientation. When

\det(A) > 0

, the pair

(\mathbf{u}, \mathbf{v})

follows the standard counterclockwise ordering — rotating from

\mathbf{u}

toward

\mathbf{v}

sweeps counterclockwise. When

\det(A) < 0

, the pair is clockwise. When

\det(A) = 0

, the two vectors are parallel and the parallelogram degenerates to a line segment with zero area.

Example

For

A = \begin{pmatrix} 3 & 1 \\ 0 & 4 \end{pmatrix}

, the columns are

(3, 0)

and

(1, 4)

. The determinant is

3 \cdot 4 - 1 \cdot 0 = 12

. The parallelogram has area

12

, and since the sign is positive, the columns are counterclockwise-ordered.

For

A = \begin{pmatrix} 1 & 3 \\ 4 & 0 \end{pmatrix}

, the determinant is

1 \cdot 0 - 3 \cdot 4 = -12

. The area is still

12

, but the negative sign means the columns are clockwise-ordered — swapping the columns reversed the orientation.

DownloadPNG2×PDF

What the sign says

The columns

\mathbf{u}

and

\mathbf{v}

of a

2 \times 2

matrix span a parallelogram, and

\det(A) = ad - bc

is its signed area. The sign encodes orientation.

Signed area

\det(A) = ad - bc

split by the sign of $\det(A)$

+

Case 1

\det(A) > 0

Counterclockwise

rotating from

\mathbf{u}

toward

\mathbf{v}

sweeps counterclockwise

e.g.columns

(3, 0)

,

(1, 4)

:

\det = 12

−

Case 2

\det(A) < 0

Clockwise

the pair is clockwise-ordered

e.g.columns

(1, 4)

,

(3, 0)

:

\det = -12

0

Case 3

\det(A) = 0

Parallel

the parallelogram degenerates to a line segment with zero area

Takeaway

The size is the area, the sign is the orientation:

\det = 12

and

\det = -12

give the same area, and swapping the columns reverses the orientation.

The sign of a 2×2 determinant·/linear-algebra/determinants/geometryLearn Math Class

The unit square after the matrix has acted

The square has become a parallelogram, and the factor by which its area changed is the determinant. Every other region on the plane is scaled by that same factor, however it is shaped, which is what lets one number describe the whole map. Change the matrix and watch the parallelogram follow on the linear transformation explorer.

The sign of that number, set aside for now, is what the orientation section below is about.

Determinant Geometry Notation

Notation

Determinant Geometry Notation

The same three marks appear here as in the algebraic treatment, but each is read for a different thing: the sign becomes a direction, the bars become a size, and a matrix of derivatives inherits the whole apparatus so it can measure how a change of coordinates stretches space. All catalogued among the linear algebra symbols.

\det(A)

, the

|A|

bars and the

\Delta

tradition come from determinant notation;

\|\mathbf{v}\|

from norm notation; the cross product and its determinant mnemonic from cross product notation;

T

and its matrix from matrix representation notation.

\det(A) < 0

a negative determinant — the map reverses orientation

Here the minus sign is geometric information, not an error: it records that the transformation flips handedness, turning a counterclockwise pair clockwise. Orientation below reads it directly; the algebraic treatment only notes that determinants can be negative, without saying what the sign means.

CasesA reflection has

\det = -1

, a rotation

\det = +1

, and

\det = 0

is the degenerate case where the image collapses to lower dimension — three readings of one number, and the reason "signed area" carries its adjective.

Do not confuseReading the sign as a mistake to be dropped. Taking

|\det|

too early discards the orientation the sign was carrying — the absolute value is the right move only when a size is wanted, which is exactly the next entry's distinction.

|\det(A)|

the absolute value of the determinant — the area or volume scale factor

A double-decker of bars with two different jobs: the inner pair (or the

\det

spelling) makes a number from a matrix, the outer pair strips its sign to leave a magnitude. That magnitude is the factor by which areas in

\mathbb{R}^2

or volumes in

\mathbb{R}^3

are multiplied.

CasesThe nesting is why this page prefers

|\det(A)|

over

||A||

— writing the determinant with its own word keeps the two bar-pairs visually distinct, and norm notation has already claimed the doubled bar for something else entirely.

Do not confuseThe norm of a matrix.

\|A\|

measures how far the matrix moves vectors;

|\det(A)|

measures how much it inflates volume. Both are "size", and they are different numbers with different units.

J_T(\mathbf{x})

,

\det J_T

the Jacobian matrix at x, and the Jacobian determinant

The local version of everything above:

J_T

collects the partial derivatives

\frac{\partial T_i}{\partial x_j}

into a matrix, so

\det J_T

is the volume scale factor at a point for a map that need not be linear — The Change-of-Variables Formula below is where it earns its keep.

Also written

\frac{\partial(x, y)}{\partial(u, v)}

— the fraction-of-tuples spelling, standard in multivariable calculus and integral change-of-variables, where the numerator lists the output coordinates and the denominator the input ones. It denotes the determinant, not the matrix.

Do not confuseMatrix and determinant sharing one word. "The Jacobian" means the matrix in some texts and its determinant in others; here

J_T

is the matrix and

\det J_T

the number, and an integral that asks for "the Jacobian" always wants the number.

All linear algebra symbols →↑ Determinants

Signed Volume in Three Dimensions

Three vectors in

\mathbb{R}^3

span a parallelepiped — the three-dimensional analogue of a parallelogram. Arranging these vectors as columns of a

3 \times 3

matrix

A

, the determinant equals the signed volume of that parallelepiped.

This is identical to the scalar triple product. If the columns of

A

are

\mathbf{a}

,

\mathbf{b}

,

\mathbf{c}

, then

Determinant Signed Volume 3D

\text{signed volume}(\mathbf{a}, \mathbf{b}, \mathbf{c}) = \det\!\begin{pmatrix} \mathbf{a} & \mathbf{b} & \mathbf{c} \end{pmatrix} = \mathbf{a} \cdot (\mathbf{b} \times \mathbf{c})

Determinant signed volume 3d →

The dot product projects

\mathbf{a}

onto the direction perpendicular to the base parallelogram spanned by

\mathbf{b}

and

\mathbf{c}

, and

|\mathbf{b} \times \mathbf{c}|

gives the base area. Their product is the volume.

The sign reflects handedness. When

\det(A) > 0

, the three column vectors form a right-handed system — the same orientation as the standard basis

\mathbf{e}_1, \mathbf{e}_2, \mathbf{e}_3

. When

\det(A) < 0

, they form a left-handed system. When

\det(A) = 0

, the three vectors are coplanar and the parallelepiped is flat.

Example

Let

\mathbf{a} = (1, 0, 0)

,

\mathbf{b} = (0, 2, 0)

,

\mathbf{c} = (0, 0, 3)

. The matrix

A = \begin{pmatrix} 1 & 0 & 0 \\ 0 & 2 & 0 \\ 0 & 0 & 3 \end{pmatrix}

has

\det(A) = 6

. The parallelepiped is an axis-aligned box with side lengths

1, 2, 3

and volume

6

, oriented right-handedly.

Swapping two columns — say placing

\mathbf{b}

first and

\mathbf{a}

second — gives

\det = -6

. The volume is unchanged, but the orientation flips to left-handed.

DownloadPNG2×PDF

What the sign says

For three vectors

\mathbf{a}

,

\mathbf{b}

,

\mathbf{c}

in

\mathbb{R}^3

, the determinant of the matrix with those columns is the signed volume of the parallelepiped they span. Its sign reflects handedness.

Signed volume

\det(A) = \mathbf{a} \cdot (\mathbf{b} \times \mathbf{c})

split by the sign of $\det(A)$

+

Case 1

\det(A) > 0

Right-handed

the same orientation as the standard basis

\mathbf{e}_1, \mathbf{e}_2, \mathbf{e}_3

e.g.columns

(1,0,0)

,

(0,2,0)

,

(0,0,3)

:

\det = 6

−

Case 2

\det(A) < 0

Left-handed

the opposite orientation; the volume is

|\det(A)|

e.g.the same columns with the first two swapped:

\det = -6

0

Case 3

\det(A) = 0

Flat

the three vectors are coplanar and the parallelepiped has no volume

Takeaway

Swapping two columns flips the sign but not the volume: the box with

\det = 6

becomes

\det = -6

, left-handed instead of right-handed.

The sign of the determinant in three dimensions·/linear-algebra/determinants/geometryLearn Math Class

The General Case: n-Dimensional Volume Scaling

The pattern generalizes to arbitrary dimension. For an

n \times n

matrix

A

, the absolute value

|\det(A)|

is the factor by which the linear map

x \mapsto Ax

scales

n

-dimensional volumes.

More precisely, consider the unit hypercube in

\mathbb{R}^n

— the set of all vectors whose components lie between

0

and

1

. Its

n

-dimensional volume is

1

. The image of this hypercube under the map

x \mapsto Ax

is a parallelepiped whose

n

-dimensional volume equals

|\det(A)|

.

This scaling factor applies uniformly to every region, not just the unit hypercube. If

S

is any measurable subset of

\mathbb{R}^n

with volume

V

, then the image

A(S) = \{Ax : x \in S\}

has volume

|\det(A)| \cdot V

. Formally:

Determinant Volume Scaling Factor

\operatorname{vol}(A(S)) =|\det(A)|\cdot \operatorname{vol}(S)

Determinant volume scaling factor →

Pitfall: Pulling the scaling factor out once

Taking the map

\mathbf{x} \mapsto 2\mathbf{x}

on

\mathbb{R}^3

to scale volumes by

2

treats

\det(2I)

as

2\det(I) = 2

. All three directions double, so

\det(2I) = 2^3 = 8

; in general

\det(kA) = k^n \det(A)

.

Three regimes emerge from this interpretation. When

|\det(A)| > 1

, the map expands volumes. When

0 < |\det(A)| < 1

, it compresses volumes. When

|\det(A)| = 1

, volumes are preserved — the map is volume-preserving. Rotations and reflections both fall in this last category.

When

\det(A) = 0

, the map collapses at least one dimension entirely. The image of

\mathbb{R}^n

under a singular matrix is a subspace of dimension less than

n

— a hyperplane, a line, a point, or something in between — and all

n

-dimensional volumes map to zero.

DownloadPNG2×PDF

What the size of the determinant says

For an

n \times n

matrix

A

, the map

x \mapsto Ax

scales every

n

-dimensional volume by the same factor

|\det(A)|

. Comparing that factor with

1

and with

0

gives four regimes.

Volume scaling

\operatorname{vol}(A(S)) = |\det(A)| \cdot \operatorname{vol}(S)

split by the size of $|\det(A)|$

↑

Case 1

|\det(A)| > 1

Expands volumes

every region comes out larger than it went in

↓

Case 2

0 < |\det(A)| < 1

Compresses volumes

every region comes out smaller

=

Case 3

|\det(A)| = 1

Volume-preserving

volumes are unchanged

e.g.rotations and reflections

0

Case 4

\det(A) = 0

Collapses a dimension

the image is a subspace of dimension less than

n

, and all

n

-dimensional volumes map to zero

e.g.a hyperplane, a line or a point

Takeaway

The absolute value discards the sign: a rotation and a reflection both have

|\det(A)| = 1

. The sign itself, orientation, is read separately.

The size of the determinant: four regimes·/linear-algebra/determinants/geometryLearn Math Class

A linear map multiplies every volume by the same factor,

|\det(A)|

, whatever the shape or size of the region.

Orientation

The sign of the determinant carries information that the absolute value discards.

RULE

A positive determinant means the linear map preserves the orientation of

\mathbb{R}^n

; a negative determinant means it reverses orientation.

Fact: A left hand turns right through 4D

No rotation of space turns a left glove into a right one, since every rotation has determinant

+1

and a mirror has

-1

. With a fourth axis it works: the half-turn

\operatorname{diag}(-1, 1, 1, -1)

in the plane of axes

1

and

4

has determinant

+1

, yet on ordinary space it acts as the mirror

x_1 \mapsto -x_1

.

In two dimensions, orientation is the sense of rotation. A positively oriented pair of vectors goes counterclockwise; a negatively oriented pair goes clockwise. A linear map with

\det(A) > 0

sends every counterclockwise pair to another counterclockwise pair. A map with

\det(A) < 0

flips counterclockwise to clockwise.

In three dimensions, orientation is handedness. The standard basis is right-handed. A transformation with positive determinant sends right-handed triples to right-handed triples; a transformation with negative determinant sends them to left-handed triples.

Standard examples make the distinction concrete. A rotation matrix in

\mathbb{R}^2

or

\mathbb{R}^3

always has

\det = +1

— it preserves both volume and orientation. A reflection across a line (in

\mathbb{R}^2

) or a plane (in

\mathbb{R}^3

) has

\det = -1

— it preserves volume but flips orientation. An improper rotation, which composes a rotation with a reflection, also has

\det = -1

.

Pitfall: Reading a determinant of 1 as a rotation

Concluding that

A

is a rotation because

\det(A) = 1

fails for the shear

\begin{pmatrix} 1 & 1 \\ 0 & 1 \end{pmatrix}

: its determinant is

1

, yet it sends

(0, 1)

to

(1, 1)

, of length

\sqrt{2}

. A determinant of

1

only says that area and orientation are kept.

The examples above start from a known map and then compute its determinant. Classifying an unknown map runs the other way: the matrix is given, its determinant is computed, and the number is then read for what it says about the map. The size of that number and its sign are the two readings that this section and the previous one provide.

When

\det(A) = 0

, orientation is undefined. The map crushes the space down to a lower-dimensional image, and the concept of clockwise versus counterclockwise or right-handed versus left-handed no longer applies in the collapsed image.

A reflection: area kept, orientation reversed

Reflection across the x-axis, the matrix with rows (1, 0) and (0, −1), whose determinant is −1. The unit square comes out the same size — |det| = 1, so no area is gained or lost — but flipped over: the turn from the first basis vector to the second, counter-clockwise before, is clockwise after. That reversal is exactly what the negative sign records. Try other mirror lines on the reflection visualizer.

No rotation or stretch can undo a reflection's flip; only another orientation-reversing map can.

Transformation	det(A)	Volume effect	Orientation effect
Rotation	+1	preserved	preserved
Reflection (across line in ℝ², plane in ℝ³)	−1	preserved	reversed
Improper rotation (rotation ∘ reflection)	−1	preserved	reversed
Singular (e.g., projection)	0	collapsed to lower dimension	undefined

Linear Transformations as Geometric Mappings

Every invertible

n \times n

matrix defines a bijective linear transformation from

\mathbb{R}^n

to itself. The determinant captures the two essential geometric facts about this transformation in a single number: it scales all volumes by

|\det(A)|

, and it either preserves or reverses orientation according to the sign.

The multiplicative property

\det(AB) = \det(A)\det(B)

takes on a clean geometric reading in this context.

RULE

Composing two linear maps multiplies their volume-scaling factors: if

A

scales volumes by

|\det(A)|

and

B

scales by

|\det(B)|

, then

AB

scales by

|\det(A)| \cdot |\det(B)|

.

The orientation effects compose as well — two orientation-reversing maps compose to an orientation-preserving one, matching the sign rule

(-1)(-1) = +1

.

The inverse

A^{-1}

undoes both effects. Its determinant is

\det(A^{-1}) = 1/\det(A)

, so it scales volumes by the reciprocal factor and restores whatever orientation

A

may have flipped.

A singular matrix (

\det = 0

) is a transformation that is not bijective. It maps

\mathbb{R}^n

onto a proper subspace, losing information in the process. No inverse exists because the collapsed dimensions cannot be recovered.

The plane flattened onto a line

The two columns now point along the same line, the grid has collapsed, and every region on the plane has been crushed to zero area. A zero determinant is exactly this: not a small scaling but a loss of a dimension, which is why nothing can be recovered and no inverse exists. Push a matrix into and out of this state on the linear transformation explorer.

Invertibility, independence of the columns and a non-zero determinant are three descriptions of this single picture not collapsing.

The Change-of-Variables Formula

The geometric interpretation of the determinant extends beyond linear maps through the change-of-variables formula in multivariable integration. Given a continuously differentiable transformation

T: \mathbb{R}^n \to \mathbb{R}^n

that is one-to-one on the region of integration, with

\det(J_T) \neq 0

there, the formula reads

The Change-of-Variables Formula

\int f(\mathbf{y}) \, d\mathbf{y} = \int f(T(\mathbf{x})) \, |\det(J_T(\mathbf{x}))| \, d\mathbf{x}

where

J_T(\mathbf{x})

is the Jacobian matrix of

T

at the point

\mathbf{x}

— the

n \times n

matrix of all partial derivatives

\frac{\partial T_i}{\partial x_j}

.

Pitfall: Dropping the Jacobian factor

Writing

dA = dr\,d\theta

in polar coordinates gives the unit disk an area of

\int_0^{2\pi}\!\int_0^1 dr\,d\theta = 2\pi

. With the factor

|\det(J_T)| = r

, the area is

\int_0^{2\pi}\!\int_0^1 r\,dr\,d\theta = \pi

.

The Jacobian matrix is the best linear approximation to

T

near

\mathbf{x}

. At each point, it acts like a linear transformation on a small neighborhood, and its determinant measures the local volume distortion. The absolute value appears because the integral accumulates unsigned volume — areas and volumes are always non-negative regardless of orientation.

In single-variable calculus, the substitution rule

\int f(g(x)) \, g'(x) \, dx

uses the derivative

g'(x)

as a one-dimensional scaling factor. The Jacobian determinant is the natural generalization:

g'(x)

is a

1 \times 1

"determinant," and

|\det(J_T)|

plays the same role in

n

dimensions.

The most familiar instances are polar coordinates in

\mathbb{R}^2

, where the Jacobian determinant is

r

, and spherical coordinates in

\mathbb{R}^3

, where it is

r^2 \sin\phi

. Both factors arise from computing

\det(J_T)

for the respective coordinate transformations.

The Jacobian determinant is the local area or volume factor, and every change of variables in an integral must include it.

Area and Volume Formulas from Determinants

The geometric interpretation yields direct coordinate-based formulas for the areas and volumes of common shapes.

Triangle Area

Given three vertices

(x_1, y_1)

,

(x_2, y_2)

,

(x_3, y_3)

in the plane, the area of the triangle they form is

Triangle Area via Determinant

\text{Area} = \frac{1}{2} \left|\det\!\begin{pmatrix} x_2 - x_1 & x_3 - x_1 \\ y_2 - y_1 & y_3 - y_1 \end{pmatrix} \right|

Triangle area via determinant →

The

2 \times 2

matrix has as its columns the edge vectors from vertex

1

to the other two vertices. The determinant gives the signed area of the parallelogram spanned by these edges, and the triangle occupies exactly half of that parallelogram.

For the vertices

(1, 2)

,

(4, 6)

,

(3, 1)

: the edge vectors are

(3, 4)

and

(2, -1)

, the determinant is

3(-1) - 4(2) = -11

, and the area is

\frac{11}{2}

.

Pitfall: Using the vertices instead of edge vectors

Putting the vertices

(4, 6)

and

(3, 1)

in as columns gives

\frac{1}{2}|4 \cdot 1 - 3 \cdot 6| = 7

, the area of the triangle with corners

(0, 0)

,

(4, 6)

,

(3, 1)

. The edge vectors

(3, 4)

and

(2, -1)

give the true area

\frac{11}{2}

.

Each edge vector is a difference of two vertices, so the formula depends only on the shape of the triangle and not on where it sits in the plane: moving all three vertices by the same amount leaves both edge vectors unchanged. Any vertex can serve as vertex

1

. Another choice gives different edge vectors and may flip the sign of the determinant, but its absolute value, and so the area, stays the same.

Parallelogram Area

The parallelogram spanned by vectors

\mathbf{u}

and

\mathbf{v}

in

\mathbb{R}^2

has area

|\det(\mathbf{u} \; \mathbf{v})|

— the full absolute value of the determinant, without the factor of

\frac{1}{2}

.

Tetrahedron Volume

Given four vertices in

\mathbb{R}^3

, pick one as the origin and form the three edge vectors to the other three. The volume of the tetrahedron is

Tetrahedron Volume via Determinant

V = \frac{1}{6} \left|\det\!\begin{pmatrix} \mathbf{e}_1 & \mathbf{e}_2 & \mathbf{e}_3 \end{pmatrix} \right|

Tetrahedron volume via determinant →

Fact: Empty of grid points, yet any volume

A triangle whose only points with whole-number coordinates are its three corners always has area

\frac{1}{2}

. In space this fails: the tetrahedron with corners

(0,0,0)

,

(1,0,0)

,

(0,1,0)

,

(1,1,r)

holds no other such point, yet its determinant is

r

and its volume

\frac{r}{6}

, as large as wanted.

where

\mathbf{e}_1, \mathbf{e}_2, \mathbf{e}_3

are the edge vectors arranged as columns. The factor

\frac{1}{6}

arises because a tetrahedron occupies one-sixth of the parallelepiped spanned by its three edges.

For vertices at

(0,0,0)

,

(1,0,0)

,

(0,2,0)

,

(0,0,3)

: the edge vectors are the standard-scaled basis vectors, the determinant is

6

, and the tetrahedron volume is

1

.

When all the vertices have whole-number coordinates, every edge vector has whole-number entries, so the determinant is a whole number and the volume is a multiple of

\frac{1}{6}

. One dimension down the same holds with

\frac{1}{2}

: a triangle with whole-number corners has an area that is a multiple of

\frac{1}{2}

, as the area

\frac{11}{2}

found above shows.

Parallelepiped Volume

The parallelepiped spanned by three vectors in

\mathbb{R}^3

has volume equal to the full absolute value of the

3 \times 3

determinant formed by those vectors as columns — no fractional factor needed.

Shape	Formula	Matrix columns
Parallelogram (2D)	\|det(A)\|	the two spanning vectors u, v
Triangle (2D)	½ · \|det(A)\|	edge vectors from one vertex to the other two
Parallelepiped (3D)	\|det(A)\|	the three spanning vectors a, b, c
Tetrahedron (3D)	⅙ · \|det(A)\|	edge vectors from one vertex to the other three

Summary: Geometric Meaning by Dimension

The geometric content of the determinant takes a recognizable shape in each dimension, with the underlying picture — signed n-volume of an n-parallelepiped — running through them all. The table below collects what det(A) measures, what its sign indicates, and what its vanishing represents, dimension by dimension.

Each row below says the same three things: what the magnitude measures, what the sign records, and what zero means. Only the name of the measurement changes — length, then area, then volume, then nothing in particular. The split is between the dimensions where a picture is available and the dimension where the algebra has to carry the meaning alone.

Download

PNGRaster · 2×PDFPrint quality

Determinants · geometry

What the determinant measures, by dimension

The magnitude is a volume, the sign is an orientation, and zero is a collapse. Only the word for "volume" changes as the dimension rises — which is the argument for calling the general case a volume at all.

5dimensions

Where the picture is available3

1

One dimension§ 3

length scaling; sign = direction

\det[a] = a

A segment on a line. The determinant is the number itself, and its sign says whether the direction is preserved or reversed — the degenerate case that makes the general statement believable rather than surprising.

2

Two dimensions§ 1

sign: counterclockwise

+

, clockwise

-

|\det A| =

area of the parallelogram

The columns of

A

span a parallelogram, and the determinant is its signed area. Zero means the columns are parallel and the parallelogram has collapsed to a line — which is the same statement as the columns being linearly dependent.

3

Three dimensions§ 2

sign: right-handed

+

, left-handed

-

|\det A| =

volume of the parallelepiped

Three columns span a parallelepiped. The sign distinguishes a right-handed frame from a left-handed one — the same distinction the cross product encodes, and the reason

\mathbf{a} \cdot (\mathbf{b} \times \mathbf{c})

is a determinant.

Where it is not2

4

Four dimensions and above§ 3

orientation preserved

+

or reversed

-

|\det A| =

volume of the image of the unit cube

No picture, and none needed — the algebra is unchanged. This is where the determinant stops being a description of something already visible and becomes the definition of

n

-dimensional content, which is what the change of variables formula then relies on.

5

Determinant zero§ 5

image lies in a proper subspace

\det A = 0

The collapse case in every dimension: the image has zero content because it fits inside something lower-dimensional. Note the determinant reports that a collapse happened but not by how much — a matrix flattening

\mathbb{R}^3

to a plane and one flattening it to a point both give zero. That is what rank is for.

Read down the column and the pattern is one definition stated four ways:

|\det|

is the factor by which the unit cube’s content is scaled, the sign says whether the image is reflected, and

\det = 0

says the image has no content because it fits inside a smaller subspace. Length, area and volume are not analogies here — they are the

n = 1, 2, 3

cases of the same quantity.

What the determinant measures, by dimension·/linear-algebra/determinants/geometryLearn Math Class

That the pattern continues past three dimensions is a claim rather than an observation, and it is worth being clear about which way the definition runs. In

\mathbb{R}^2

and

\mathbb{R}^3

the determinant is checked against an area and a volume that were already understood. Above that there is no prior notion of content to check against, so the determinant supplies one —

n

-dimensional volume is defined as what the determinant measures, and the change of variables formula is what makes that definition useful rather than arbitrary.

The zero case deserves separate attention because it is the one that loses information. A determinant of zero says the image fits inside a proper subspace, but says nothing about which subspace or how far the collapse went — a matrix squashing

\mathbb{R}^3

onto a plane and one squashing it onto a point both report zero. The determinant detects collapse; rank measures it.

Common Pitfalls

Four mistakes come up again and again when the determinant is used to measure areas and volumes. Each is marked with a box where it happens in the sections above. Here each is worked through in full, with the wrong and the right working side by side.

Pulling the scaling factor out once

A map that multiplies every coordinate by the same number

k

has the matrix

kI

, and scaling a map

A

in the same way gives

kA

. The factor

k

is taken out of the determinant once, so the volume factor is taken to be

k\det(A)

.

Wrong working: for the map

\mathbf{x} \mapsto 2\mathbf{x}

on

\mathbb{R}^3

,

\det(2I) = 2\det(I) = 2

, so the unit cube is expected to go to a solid of volume

2

.

Right working: every one of the three rows of

2I

carries the factor

2

.

\det(2I) = \det\begin{pmatrix} 2 & 0 & 0 \\ 0 & 2 & 0 \\ 0 & 0 & 2 \end{pmatrix} = 2 \cdot 2 \cdot 2 = 8

The unit cube goes to the cube with side

2

, whose volume is

8

. For an

n \times n

matrix,

\det(kA) = k^n \det(A)

.

Why it happens: the determinant is linear in each row separately, so a factor in one row comes out once. Multiplying the whole matrix by

k

puts the factor in all

n

rows. For a single number, the case

n = 1

, scaling by

k

really does scale by

k

, and that habit is carried over.

How to check: picture the unit square or cube. Doubling every side of a square gives area

4

, and doubling every side of a cube gives volume

8

, never

2

.

Reading a determinant of 1 as a rotation

Rotations have determinant

+1

and reflections

-1

, and the statement is read backwards: a matrix with determinant

1

is taken to be a rotation, one with

-1

a reflection. The determinant records only how volumes scale and whether orientation is kept. Lengths and angles can still change.

Wrong working: for

S = \begin{pmatrix} 1 & 1 \\ 0 & 1 \end{pmatrix}

,

\det(S) = 1 \cdot 1 - 1 \cdot 0 = 1

, so

S

is called a rotation.

Right working:

S

sends

(1, 0)

to

(1, 0)

and

(0, 1)

to

(1, 1)

. The second image has length

\sqrt{2}

, and the two images meet at

45^\circ

instead of

90^\circ

.

S

is a shear: it keeps area and orientation but changes the shape of the square. A rotation also needs

S^T S = I

, and here

S^T S = \begin{pmatrix} 1 & 1 \\ 1 & 2 \end{pmatrix} \neq I

Why it happens: the standard examples are always given in one direction, from the map to its determinant, and "rotations have determinant

1

" is remembered as "determinant

1

means rotation".

How to check: look at the columns. For a rotation or a reflection both columns have length

1

and are perpendicular. The second column of

S

is

(1, 1)

, of length

\sqrt{2}

.

Dropping the Jacobian factor

When an integral is moved to new coordinates, the differentials are replaced, for example

dx\,dy

by

dr\,d\theta

, but the factor

|\det(J_T)|

that measures how the map stretches area is left out.

Wrong working: the area of the unit disk in polar coordinates,

\int_0^{2\pi}\!\int_0^1 dr\,d\theta = 2\pi

Right working: for

T(r, \theta) = (r\cos\theta, r\sin\theta)

,

\det(J_T) = \det\begin{pmatrix} \cos\theta & -r\sin\theta \\ \sin\theta & r\cos\theta \end{pmatrix} = r\cos^2\theta + r\sin^2\theta = r

so the area is

\int_0^{2\pi}\!\int_0^1 r\,dr\,d\theta = \int_0^{2\pi} \frac{1}{2}\,d\theta = \pi

Why it happens: in the new variables the region is a plain rectangle,

0 \le r \le 1

and

0 \le \theta \le 2\pi

, and the integral looks complete without the factor. The stretching done by the map cannot be seen in the new coordinates.

How to check: integrate

1

over a region whose area is known. The unit disk must give

\pi

; the wrong answer

2\pi

is the area of the rectangle in the

(r, \theta)

plane, not of the disk.

Using the vertices instead of edge vectors

The two columns of the

2 \times 2

matrix are filled with two of the vertices themselves instead of the edge vectors from vertex

1

. This measures a different triangle, the one with a corner at the origin.

Wrong working: for the vertices

(1, 2)

,

(4, 6)

,

(3, 1)

, taking

(4, 6)

and

(3, 1)

as columns,

\frac{1}{2}\left|\det\begin{pmatrix} 4 & 3 \\ 6 & 1 \end{pmatrix}\right| = \frac{1}{2}|4 - 18| = 7

Right working: the edge vectors are

(4 - 1, 6 - 2) = (3, 4)

and

(3 - 1, 1 - 2) = (2, -1)

, and

\frac{1}{2}\left|\det\begin{pmatrix} 3 & 2 \\ 4 & -1 \end{pmatrix}\right| = \frac{1}{2}|-3 - 8| = \frac{11}{2}

Why it happens: the parallelogram formula takes vectors that start at the origin, and vertex coordinates look like such vectors. The shortcut is right only when one of the vertices is the origin.

How to check: shift all three vertices by the same amount; the area must not change. Moving

(1, 2)

to the origin gives the vertices

(0, 0)

,

(3, 4)

,

(2, -1)

, and the shortcut now gives

\frac{11}{2}

instead of

7

, so it cannot have been measuring the original triangle.

Determinant Geometry FAQ

Should you drop the minus sign from a negative determinant?

+

Not straight away, because the sign is carrying orientation. A rotation has determinant

+1

, a reflection

-1

, and

0

marks the degenerate case where the image collapses into a lower dimension. Taking the absolute value is the right move only once you genuinely want a size, and doing it earlier discards information.Read more →

Is $|\det(A)|$ the same as the matrix norm $\|A\|$ ?

+

No, they are different numbers measuring different things. The norm measures how far the matrix moves vectors, while

|\det(A)|

measures how much it inflates volume. Both get described loosely as the size of a matrix, which is why keeping the determinant's own word visible avoids confusing two bar notations.Read more →

Does "the Jacobian" mean the matrix or the determinant?

+

Both, depending on the text, which is a genuine source of confusion. Some authors use the word for the matrix of partial derivatives and others for its determinant. The practical rule: an integral asking for the Jacobian always wants the number, since it is the volume scaling factor that the change of variables requires.Read more →

Determinant as Area and Volume

Area, Volume, and Orientation

Signed Area in Two Dimensions

Determinant Geometry Notation

Determinant Geometry Notation

Signed Volume in Three Dimensions

The General Case: n-Dimensional Volume Scaling

Orientation

Linear Transformations as Geometric Mappings

The Change-of-Variables Formula

Area and Volume Formulas from Determinants

Triangle Area

Parallelogram Area

Tetrahedron Volume

Parallelepiped Volume

Summary: Geometric Meaning by Dimension

What the determinant measures, by dimension

Common Pitfalls

Determinant Geometry FAQ

Should you drop the minus sign from a negative determinant?

Is $|\det(A)|$ the same as the matrix norm $\|A\|$ ?

Does "the Jacobian" mean the matrix or the determinant?

Determinant as Area and Volume

Area, Volume, and Orientation

Signed Area in Two Dimensions

Determinant Geometry Notation

Determinant Geometry Notation

Signed Volume in Three Dimensions

The General Case: n-Dimensional Volume Scaling

Orientation

Linear Transformations as Geometric Mappings

The Change-of-Variables Formula

Area and Volume Formulas from Determinants

Triangle Area

Parallelogram Area

Tetrahedron Volume

Parallelepiped Volume

Summary: Geometric Meaning by Dimension

What the determinant measures, by dimension

Common Pitfalls

Determinant Geometry FAQ

Should you drop the minus sign from a negative determinant?

Is ∣det⁡(A)∣|\det(A)|∣det(A)∣ the same as the matrix norm ∥A∥\|A\|∥A∥?

Does "the Jacobian" mean the matrix or the determinant?

Is $|\det(A)|$ the same as the matrix norm $\|A\|$ ?