A subspace is a subset of a vector space that is itself a vector
space. The simplest example is a line through the origin in
the plane. For the line is definitely a subset and if we add
any two vectors on the line we remain on the line and if we
multiply any vector on the line by a scalar we remain on the
line. The same could be said for a line or plane through the
origin in 3 space. As we shall be travelling in spaces with
many many dimensions it pays to have a general definition.
- Definition 1: A subset SS of a vector
space VV is a subspace of
VV when
1.
if
xx and
yy belong to
SS then so does
x+y
x
y
.
2.
if xx belongs
to SS and
tt is real then
tx
t
x
belong to
SS.
As these are oftentimes unwieldy objects it pays to look for a
handful of vectors from which the entire subset may be generated.
For example, the set of
xx for which
x
1
+
x
2
+
x
3
+
x
4
=0
x
1
x
2
x
3
x
4
0
constitutes a subspace of
ℝ
4
ℝ
4
.
Can you 'see' this set? Do you 'see' that
-1-100
-1-100
and
-1010
-1010
and
-1001
-1001
not only belong to a set but in fact generate all
possible elements? More precisely, we say that these vectors
span the subspace of all possible solutions.
- Definition 2: Span
A finite collection
s
1
s
2
…
s
n
s
1
s
2
…
s
n
of vectors in the subspace
SS is said
to span SS
if each element of SS can be written as
a linear combination of these vectors. That is, if for
each
s∈S
s
S
there exist
nn reals
x
1
x
2
…
x
n
x
1
x
2
…
x
n
such that
s=
x
1
s
1
+
x
2
s
2
+…+
x
n
s
n
s
x
1
s
1
x
2
s
2
…
x
n
s
n
.
When attempting to generate a subspace as the span of a
handful of vectors it is natural to ask what is the fewest
number possible. The notion of linear independence helps us
clarify this issue.
- Definition 3: Linear Independence
A finite collection
s
1
s
2
…
s
n
s
1
s
2
…
s
n
of vectors is said to
be linearly independent when the only reals,
x
1
x
2
…
x
n
x
1
x
2
…
x
n
for which
x
1
+
x
2
+…+
x
n
=0
x
1
x
2
…
x
n
0
are
x
1
=
x
2
=…=
x
n
=0
x
1
x
2
…
x
n
0
.
In other words, when the
null space
of the matrix whose columns are
s
1
s
2
…
s
n
s
1
s
2
…
s
n
contains only the zero vector.
Combining these definitions, we arrive at the precise notion of a
'generating set.'
- Definition 4: Basis
Any linearly independent spanning set of a subspace
SS is called
a basis of
SS.
Though a subspace may have many bases they all have one thing
in common:
- Definition 5: Dimension
The dimension of a subspace
is the number of elements in its basis.