1. (3 pts) Let B={b1,b2} and C={c1,c2} be bases for R2, and suppose b1=4c1−c2 and b2=−6c1+c2. Write the change-of-coordinates matrix C←BP, and use it to find [x]C when [x]B=(3,−2).
2. (3 pts) Let B={(1,2),(1,3)} and C={(1,1),(0,1)} be bases for R2. Compute C←BP=PC−1PB.
3. (3 pts) In P2, let B={1,t,t2} and C={1,1+t,1+t+t2}. Find C←BP, and use it to convert [p]B=(2,3,−1) to [p]C.
4. (3 pts) Using the matrix C←BP from Problem 3, find B←CP, and verify that the two matrices are inverses.
5. (2 pts) Explain why the change-of-coordinates matrix C←BP between two bases of Rn is always invertible, and identify its inverse.
6. (3 pts) State whether each of the following is true or false, and justify each answer.
(a) The columns of C←BP are the C-coordinate vectors of the vectors in B.
(b) C←BP B←CP=I.
(c) If B=C, then C←BP=I.
(d) For the standard basis E of Rn, E←BP is the matrix whose columns are the vectors of B.
7. (3 pts) Let B and C be bases for Rn. Prove that [x]C=C←BP[x]B for all x, where the j-th column of C←BP is [bj]C.
8. (4 pts ••) Let B, C, and D be bases for Rn. Prove that D←BP=D←CP C←BP.