The rank-nullity theorem states that the dimension of the domain of a linear function is equal to the sum of the dimensions of its range (i.e., the set of values in the codomain that the function actually takes) and kernel (i.e., the set of values in the domain that are mapped to the zero vector in the codomain).
Table of contents
First of all, we need to define a linear
map
between two vector spaces
and
,
that is, a function such
that
for
any two vectors
and any two scalars
and
.
The set
is called the domain of
.
Being a vector space, the domain
has a dimension,
denoted
by
which
is equal to the number of elements of a basis of
(any basis; remember that all the bases of a space have the same number of
elements).
The range (or image) of
is the subset of the codomain
that comprises all the values taken by
as
varies over
:
As we proved in the lecture on the
range of a linear map,
is a subspace of
.
Its dimension is called rank and is denoted
by
The null space (or kernel) of
is the subset of the domain
that comprises all the values mapped by
into the zero vector of
:
As we proved in the lecture on the
null space of a linear
map,
is a subspace of
.
Its dimension is called nullity and is denoted
by
Here comes one of the most important theorems in linear algebra, called rank-nullity theorem.
Proposition
Let
and
be two linear spaces. Let
be a linear map. If
is finite, then
Let
and
choose a basis
for
.
Let
and
choose a basis
for
.
By the basis extension
theorem, we can form a new basis for
that contains
:
possibly
after re-numbering the vectors of the basis
.
Any
can be uniquely written as a linear combination of the basis
above:
where
are scalars. Then, we
have
where
in step
we have used the fact that
belong to the null space of
and, as a consequence,
.
The equation just derived holds for any
,
which implies that
is spanned by the
vectors
We
need to prove that these vectors are linearly independent. By contradiction,
suppose they are not. Then, they can be linearly combined so as to obtain the
zero
vector:
where
are the coefficients of the linear combination (not all equal to zero). By the
linearity of
,
we have
that
which
means that the
vector
belongs
to
.
Since
is a basis for
,
we can write the latter vector as a linear combination of
with coefficients
:
or
Thus,
there is a linear combination of the basis of
,
with coefficients not all equal to zero, that gives the zero vector as a
result. This is impossible because the elements of a basis are linearly
independent. Therefore, we have arrived at a contradiction, which means that
the
vectors
must
be linearly independent. We have already proved that they span
.
So, we conclude that they are a basis for
.
Hence,
Below you can find some exercises with explained solutions.
As we demonstrated in the lecture on
linear maps, a linear map
is completely specified by the values taken by
on a basis for
.
Let
be a basis for
and
be a basis for
.
Specify the function
as
follows:
Find the dimensions of
and
and then use the rank-nullity theorem to find the nullity of
.
Since there are three vectors in the
basis
,
we have
that
Any
vector
can be written as a linear
combination
where
are scalars. The transformed
vectors
can
be written as linear combinations of
and
.
As a consequence,
is a basis for
.
Therefore,
To
conclude, we
get
Please cite as:
Taboga, Marco (2021). "The rank-nullity theorem", Lectures on matrix algebra. https://www.statlect.com/matrix-algebra/rank-nullity-theorem.
Most of the learning materials found on this website are now available in a traditional textbook format.