Linear Algebra/Subspaces and Spanning sets/Solutions

SolutionsEdit

This exercise is recommended for all readers.
Problem 1

Which of these subsets of the vector space of   matrices are subspaces under the inherited operations? For each one that is a subspace, parametrize its description. For each that is not, give a condition that fails.

  1.  
  2.  
  3.  
  4.  
Answer

By Lemma 2.9, to see if each subset of   is a subspace, we need only check if it is nonempty and closed.

  1. Yes, it is easily checked to be nonempty and closed. This is a parametrization.
     
    By the way, the parametrization also shows that it is a subspace, it is given as the span of the two-matrix set, and any span is a subspace.
  2. Yes; it is easily checked to be nonempty and closed. Alternatively, as mentioned in the prior answer, the existence of a parametrization shows that it is a subspace. For the parametrization, the condition   can be rewritten as  . Then we have this.
     
  3. No. It is not closed under addition. For instance,
     
    is not in the set. (This set is also not closed under scalar multiplication, for instance, it does not contain the zero matrix.)
  4. Yes.
     
This exercise is recommended for all readers.
Problem 2

Is this a subspace of  :  ? If it is then parametrize its description.

Answer

No, it is not closed. In particular, it is not closed under scalar multiplication because it does not contain the zero polynomial.

This exercise is recommended for all readers.
Problem 3

Decide if the vector lies in the span of the set, inside of the space.

  1.  ,  , in  
  2.  ,  , in  
  3.  ,  , in  
Answer
  1. Yes, solving the linear system arising from
     
      and  .
  2. Yes; the linear system arising from  
     
    gives that  .
  3. No; any combination of the two given matrices has a zero in the upper right.
Problem 4

Which of these are members of the span   in the vector space of real-valued functions of one real variable?

  1.  
  2.  
  3.  
  4.  
Answer
  1. Yes; it is in that span since  .
  2. No, since   has no scalar solutions that work for all  . For instance, setting   to be   and   gives the two equations   and  , which are not consistent with each other.
  3. No; consider what happens on setting   to be   and  .
  4. Yes,  .
This exercise is recommended for all readers.
Problem 5

Which of these sets spans  ? That is, which of these sets has the property that any three-tall vector can be expressed as a suitable linear combination of the set's elements?

  1.  
  2.  
  3.  
  4.  
  5.  
Answer
  1. Yes, for any   this equation
     
    has the solution  ,  , and  .
  2. Yes, the equation
     
    gives rise to this
     
    so that, given any  ,  , and  , we can compute that  ,  , and  .
  3. No. In particular, the vector
     
    cannot be gotten as a linear combination since the two given vectors both have a third component of zero.
  4. Yes. The equation
     
    leads to this reduction.
     
    We have infinitely many solutions. We can, for example, set   to be zero and solve for  ,  , and   in terms of  ,  , and   by the usual methods of back-substitution.
  5. No. The equation
     
    leads to this reduction.
     
    This shows that not every three-tall vector can be so expressed. Only the vectors satisfying the restriction that   are in the span. (To see that any such vector is indeed expressible, take   and   to be zero and solve for   and   in terms of  ,  , and   by back-substitution.)
This exercise is recommended for all readers.
Problem 6

Parametrize each subspace's description. Then express each subspace as a span.

  1. The subset   of the three-wide row vectors
  2. This subset of  
     
  3. This subset of  
     
  4. The subset   of  
  5. The subset of   of quadratic polynomials   such that  
Answer
  1.   The obvious choice for the set that spans is  .
  2.   One set that spans this space consists of those three matrices.
  3. The system
     
    gives   and  . So one description is this.
     
    That shows that a set spanning this subspace consists of those two matrices.
  4. The   gives  . So the subspace is the span of the set  .
  5. The set   parametrized as   has the spanning set  .
This exercise is recommended for all readers.
Problem 7

Find a set to span the given subspace of the given space. (Hint. Parametrize each.)

  1. the  -plane in  
  2.   in  
  3.   in  
  4.   in  
  5. The set   in the space  
  6.   in  
Answer

Each answer given is only one out of many possible.

  1. We can parametrize in this way
     
    giving this for a spanning set.
     
  2. As a way to solve it, is to express   as   to get this parametrization.
     
    and get this as a spanning set.
     
  3. Here as a way to go, is to notice that   and   depend on   and  , but not each other, suggesting to choose   and   as free variables. Then re-expressing,   and   to obtain this parametrization.
     
    and for a possible spanning set we have.
     
  4. Again, as a way to do it, re-express   and   to get parametrization.
     
    which gives this for a spanning set.
     
  5. As there are no restrictions, we can get a spaning set right away, skipping parametrization.
     
  6. For same reason as the previous subproblem, we can skip parametrization and go to a spaning set immediatedly.
     
Problem 8

Is   a subspace of  ?

Answer

Technically, no. Subspaces of   are sets of three-tall vectors, while   is a set of two-tall vectors. Clearly though,   is "just like" this subspace of  .

 
This exercise is recommended for all readers.
Problem 9

Decide if each is a subspace of the vector space of real-valued functions of one real variable.

  1. The even functions  . For example, two members of this set are   and  .
  2. The odd functions  . Two members are   and  .
Answer

Of course, the addition and scalar multiplication operations are the ones inherited from the enclosing space.

  1. This is a subspace. It is not empty as it contains at least the two example functions given. It is closed because if   are even and   are scalars then we have this.
     
  2. This is also a subspace; the check is similar to the prior one.
Problem 10

Example 2.16 says that for any vector   that is an element of a vector space  , the set   is a subspace of  . (This is of course, simply the span of the singleton set  .) Must any such subspace be a proper subspace, or can it be improper?

Answer

It can be improper. If   then this is a trivial subspace. At the opposite extreme, if the vector space is   and   then the subspace is all of  .

Problem 11

An example following the definition of a vector space shows that the solution set of a homogeneous linear system is a vector space. In the terminology of this subsection, it is a subspace of   where the system has   variables. What about a non-homogeneous linear system; do its solutions form a subspace (under the inherited operations)?

Answer

No, such a set is not closed. For one thing, it does not contain the zero vector.

Problem 12

Example 2.19 shows that   has infinitely many subspaces. Does every nontrivial space have infinitely many subspaces?

Answer

No. The only subspaces of   are the space itself and its trivial subspace. Any subspace   of   that contains a nonzero member   must contain the set of all of its scalar multiples  . But this set is all of  .

Problem 13

Finish the proof of Lemma 2.9.

Answer

Item (1) is checked in the text.

Item (2) has five conditions. First, for closure, if   and   then   as  . Second, because the operations in   are inherited from  , for   and  , the scalar product   in   equals the product   in  , and that equals   in  , which equals   in  .

The check for the third, fourth, and fifth conditions are similar to the second conditions's check just given.

Problem 14

Show that each vector space has only one trivial subspace.

Answer

An exercise in the prior subsection shows that every vector space has only one zero vector (that is, there is only one vector that is the additive identity element of the space). But a trivial space has only one element and that element must be this (unique) zero vector.

This exercise is recommended for all readers.
Problem 15

Show that for any subset   of a vector space, the span of the span equals the span  . (Hint. Members of   are linear combinations of members of  . Members of   are linear combinations of linear combinations of members of  .)

Answer

As the hint suggests, the basic reason is the Linear Combination Lemma from the first chapter. For the full proof, we will show mutual containment between the two sets.

The first containment   is an instance of the more general, and obvious, fact that for any subset   of a vector space,  .

For the other containment, that  , take   vectors from  , namely  , ...,  , and note that any linear combination of those

 

is a linear combination of elements of  

 

and so is in  . That is, simply recall that a linear combination of linear combinations (of members of  ) is a linear combination (again of members of  ).

Problem 16

All of the subspaces that we've seen use zero in their description in some way. For example, the subspace in Example 2.3 consists of all the vectors from   with a second component of zero. In contrast, the collection of vectors from   with a second component of one does not form a subspace (it is not closed under scalar multiplication). Another example is Example 2.2, where the condition on the vectors is that the three components add to zero. If the condition were that the three components add to one then it would not be a subspace (again, it would fail to be closed). This exercise shows that a reliance on zero is not strictly necessary. Consider the set

 

under these operations.

 
  1. Show that it is not a subspace of  . (Hint. See Example 2.5).
  2. Show that it is a vector space. Note that by the prior item, Lemma 2.9 can not apply.
  3. Show that any subspace of   must pass through the origin, and so any subspace of   must involve zero in its description. Does the converse hold? Does any subset of   that contains the origin become a subspace when given the inherited operations?
Answer
  1. It is not a subspace because these are not the inherited operations. For one thing, in this space,
     
    while this does not, of course, hold in  .
  2. We can combine the argument showing closure under addition with the argument showing closure under scalar multiplication into one single argument showing closure under linear combinations of two vectors. If   are in   then
     
    (note that the definition of addition in this space is that the first components combine as  , so the first component of the last vector does not say " "). Adding the three components of the last vector gives  . Most of the other checks of the conditions are easy (although the oddness of the operations keeps them from being routine). Commutativity of addition goes like this.
     
    Associativity of addition has
     
    while
     
    and they are equal. The identity element with respect to this addition operation works this way
     
    and the additive inverse is similar.
     
    The conditions on scalar multiplication are also easy. For the first condition,
     
    while
     
    and the two are equal. The second condition compares
     
    with
     
    and they are equal. For the third condition,
     
    while
     
    and the two are equal. For scalar multiplication by   we have this.
     
    Thus all the conditions on a vector space are met by these two operations. Remark. A way to understand this vector space is to think of it as the plane in  
     
    displaced away from the origin by   along the  -axis. Then addition becomes: to add two members of this space,
     
    (such that   and  ) move them back by   to place them in   and add as usual,
     
    and then move the result back out by   along the  -axis.
     
    Scalar multiplication is similar.
  3. For the subspace to be closed under the inherited scalar multiplication, where   is a member of that subspace,
     
    must also be a member. The converse does not hold. Here is a subset of   that contains the origin
     
    (this subset has only two elements) but is not a subspace.
Problem 17

We can give a justification for the convention that the sum of zero-many vectors equals the zero vector. Consider this sum of three vectors  .

  1. What is the difference between this sum of three vectors and the sum of the first two of these three?
  2. What is the difference between the prior sum and the sum of just the first one vector?
  3. What should be the difference between the prior sum of one vector and the sum of no vectors?
  4. So what should be the definition of the sum of no vectors?
Answer
  1.  
  2.  
  3. Surely,  .
  4. Taking the one-long sum and subtracting gives ( .
Problem 18

Is a space determined by its subspaces? That is, if two vector spaces have the same subspaces, must the two be equal?

Answer

Yes; any space is a subspace of itself, so each space contains the other.

Problem 19
  1. Give a set that is closed under scalar multiplication but not addition.
  2. Give a set closed under addition but not scalar multiplication.
  3. Give a set closed under neither.
Answer
  1. The union of the  -axis and the  -axis in   is one.
  2. The set of integers, as a subset of  , is one.
  3. The subset   of   is one, where   is any nonzero vector.
Problem 20

Show that the span of a set of vectors does not depend on the order in which the vectors are listed in that set.

Answer

Because vector space addition is commutative, a reordering of summands leaves a linear combination unchanged.

Problem 21

Which trivial subspace is the span of the empty set? Is it

 

or some other subspace?

Answer

We always consider that span in the context of an enclosing space.

Problem 22

Show that if a vector is in the span of a set then adding that vector to the set won't make the span any bigger. Is that also "only if"?

Answer

It is both "if" and "only if".

For "if", let   be a subset of a vector space   and assume   satisfies   where   are scalars and  . We must show that  .

Containment one way,   is obvious. For the other direction,  , note that if a vector is in the set on the left then it has the form   where the  's are scalars and the  's are in  . Rewrite that as   and note that the result is a member of the span of  .

The "only if" is clearly true— adding   enlarges the span to include at least  .

This exercise is recommended for all readers.
Problem 23

Subspaces are subsets and so we naturally consider how "is a subspace of" interacts with the usual set operations.

  1. If   are subspaces of a vector space, must   be a subspace? Always? Sometimes? Never?
  2. Must   be a subspace?
  3. If   is a subspace, must its complement be a subspace?

(Hint. Try some test subspaces from Example 2.19.)

Answer
  1. Always. Assume that   are subspaces of  . Note that their intersection is not empty as both contain the zero vector. If   and   are scalars then   because each vector is in   and so a linear combination is in  , and   for the same reason. Thus the intersection is closed. Now Lemma 2.19 applies.
  2. Sometimes (more precisely, only if   or  ). To see the answer is not "always", take   to be  , take   to be the  -axis, and   to be the  -axis. Note that
     
    as the sum is in neither   nor  . The answer is not "never" because if   or   then clearly   is a subspace. To show that   is a subspace only if one subspace contains the other, we assume that   and   and prove that the union is not a subspace. The assumption that   is not a subset of   means that there is an   with  . The other assumption gives a   with  . Consider  . Note that sum is not an element of   or else   would be in  , which it is not. Similarly the sum is not an element of  . Hence the sum is not an element of  , and so the union is not a subspace.
  3. Never. As   is a subspace, it contains the zero vector, and therefore the set that is  's complement does not. Without the zero vector, the complement cannot be a vector space.
This exercise is recommended for all readers.
Problem 24

Does the span of a set depend on the enclosing space? That is, if   is a subspace of   and   is a subset of   (and so also a subset of  ), might the span of   in   differ from the span of   in  ?

Answer

The span of a set does not depend on the enclosing space. A linear combination of vectors from   gives the same sum whether we regard the operations as those of   or as those of  , because the operations of   are inherited from  .

Problem 25

Is the relation "is a subspace of" transitive? That is, if   is a subspace of   and   is a subspace of  , must   be a subspace of  ?

Answer

It is; apply Lemma 2.19. (You must consider the following. Suppose   is a subspace of a vector space   and suppose   is a subspace. From which space does   inherit its operations? The answer is that it doesn't matter—   will inherit the same operations in either case.)

This exercise is recommended for all readers.
Problem 26

Because "span of" is an operation on sets we naturally consider how it interacts with the usual set operations.

  1. If   are subsets of a vector space, is  ? Always? Sometimes? Never?
  2. If   are subsets of a vector space, is  ?
  3. If   are subsets of a vector space, is  ?
  4. Is the span of the complement equal to the complement of the span?
Answer
  1. Always; if   then a linear combination of elements of   is also a linear combination of elements of  .
  2. Sometimes (more precisely, if and only if   or  ). The answer is not "always" as is shown by this example from  
     
    because of this.
     
    The answer is not "never" because if either set contains the other then equality is clear. We can characterize equality as happening only when either set contains the other by assuming   (implying the existence of a vector   with  ) and   (giving a   with  ), noting  , and showing that  .
  3. Sometimes. Clearly   because any linear combination of vectors from   is a combination of vectors from   and also a combination of vectors from  . Containment the other way does not always hold. For instance, in  , take
     
    so that   is the  -axis but   is the trivial subspace. Characterizing exactly when equality holds is tough. Clearly equality holds if either set contains the other, but that is not "only if" by this example in  .
     
  4. Never, as the span of the complement is a subspace, while the complement of the span is not (it does not contain the zero vector).
Problem 27

Reprove Lemma 2.15 without doing the empty set separately.

Answer

Call the subset  . By Lemma 2.9, we need to check that   is closed under linear combinations. If   then for any   we have

 

which is an element of  . (Remark. If the set   is empty, then that "if ... then ..." statement is vacuously true.)

Problem 28

Find a structure that is closed under linear combinations, and yet is not a vector space. (Remark. This is a bit of a trick question.)

Answer

For this to happen, one of the conditions giving the sensibleness of the addition and scalar multiplication operations must be violated. Consider   with these operations.

 

The set   is closed under these operations. But it is not a vector space.