Date: 4/21/2009 Linear Algebra MATH 320 This examines the connections between a few of the concepts we have learned in linear alge- bra this semester. Being very comfortable with these questions, their answers, and the ideas behind them will go a long way towards helping you do well on linear algebra questions in this class and more generally will give you a strong basis in linear algebra whenever it comes up again in your life. I will not post answers to these questions (so please don?t ask me to); if you would like to know how you are doing or if you would like to discuss the answers to these questions, Professor Smith, Sarah and I will be more than happy to meet with you. Throughout, let A be a square n n matrix. Here are ten properties that are equivalent to A being invertible. (What I mean by this is that if A has any of the ten properties below, then it has the other nine properties and it is invertible, and if A is invertible, then it has all the ten properties below. In other words, being invertible and these ten properties are all the same thing.) 1) A is row-equivalent to the n n identity matrix. 2) A has n leading entries. 3) det A6= 0. 4) The rank of A is n. 5) The equation Ax=0 only has the trivial solution x=0. 6) The equation Ax=b has exactly one solution for each b in Rn. 7) The columns of A are linearly independent. 8) The columns of A span Rn. 9) The columns of A form a basis of Rn. 10) The number 0 is not an eigenvalue of A. Determinant and Inverses If det A = 10, is A invertible? If det A = , is A invertible? If det A = 0, is A invertible? Row operations and determinant How many di erent kinds of row operations are there and how does each change the deter- minant? So overall, what are the ways in which row operations can change the determinant? 1 Can a matrix with a non-zero determinant be row equivalent to a matrix with zero deter- minant? Can a matrix with zero determinant be row equivalent to a matrix with non-zero determinant? Are all matrices with non-zero determinant row equivalent? If A is row equivalent to B and B is row equivalent to C, is A row equivalent to C? Are all matrices with determinant zero row equivalent? (They aren?t. Give an example of two 3 3 matrices that both have determinant zero and that are not row equivalent.) Column space and row space If A is row equivalent to B, do A and B have the same row space? Do A and B have the same column space? (You can nd the answer to these questions in the book.) If A is row equivalent to B, do A and B have the same rank? What is the rank of the n n identity matrix? If A is row equivalent to the identity matrix, what is the rank of A? If A has k leading entries, what is the rank of A? Linear independence, spanning and bases In R3, can 4 vectors be linearly independent? Can 3 vectors be linearly independent? Can 2 vectors be linearly independent? Can 3 vectors be linearly dependent? More generally, in Rn, how many vectors can be linearly independent, and at which point do you know for sure the vectors are linearly dependent? Can 4 vectors span all of R3? Can 3 vectors span all of R3? Can 2 vectors span all of R3? Is it possible for 3 vectors to not span all of R3? More generally, in Rn, how many vectors are necessary to span all of Rn? If I have 3 linearly independent vectors in R3, do they span R3? What can you say in the case of Rn? If 3 vectors span R3, are they linearly independent? Again, what can you say in the case of Rn? Write down the de nition of a basis. If I have 3 linearly independent vectors in R3, do they form a basis of R3? What if I 2 have 2 linearly independent vectors? If 3 vectors span R3, do they form a basis of R3? What if 4 vectors span R3, do they form a basis then? Look at facts 7), 8) and 9). Does it make sense that they are all the same now? Solutions and column spaces Let A be any n n matrix. How many solutions can the system Ax=b have, if b is any vector? How many solutions can the system Ax=0 have? If Ax=0 has in nitely many solutions, what is the determinant of A? If Ax=0 has in- nitely many solutions, is A row equivalent to the identity matrix? Suppose I am row-reducing the matrix A, and I end up with n leading entries (remem- ber A is an n n matrix). Is A invertible? What is the column space of A? What are two di erent bases you could give for the column space of A? How many solutions does the system Ax=0 have? Suppose that I am row-reducing the matrix A, and I end up with 2 rows of zeros when I am done. How many leading entries does A have? How many solutions does Ax=0 have? The vectors x such that Ax=0 form a subspace of Rn (prove this); what is its dimension? What is the dimension of the column space of A? If you had to give a basis for the column space, how many vectors would be in it? (For this set of questions, you might want to work out a few examples of matrices with two rows of zeros when they are row reduced and see what happens.) More generally, suppose that I am row-reducing the matrix A and I end up with k rows of zeros when I am done, where 0 k n. How many leading entries does A have? What is the dimension of the subspace of Rn formed by the vectors x such that Ax=0? What is the dimension of the column space of A? Eigenvalues What does it mean if 0 is an eigenvalue of A? What does it have to do with Fact 5)? 3