Вы находитесь на странице: 1из 15

1. Show that f is the function form R3 to R3 which rotates every vector A counterclockwise about some axis L.

In the book, we see if the axis L is the unit vector u = (a,b,c) then the standard matrix for the function f is a). Prove that this transformation is linear. By definition, a transformation is linear if and only if
(1). (2). f(u+v) = f(u) + f(v) f(cu) = cf(u). (1) (2)

In order to prove that these properties hold for the rotation we shall use geometry. If we take the sum of two vectors and then rotate the resulting vector, it will be the same as rotating the two vectors and then taking the sum. This is because the angle between the two vectors does not change when they are rotated separately, since they are rotated through the same angle, and their lengths do not change. The second property also applies, because if you multiply the vector by a scalar the result will be parallel to the original vector. So, when you rotate the vector A it will be parallel to the same line as the vector rotated without multiplying the vector A by a scalar. The magnitude will be the same for either method. Since magnitude is not changed during rotation, it will be the same if you multiply it by a scalar before or after the rotation. b). Show that the column-vectors in the standard matrix A of f are pairwise orthogonal and their lengths are equal to 1. We can write the standard matrix, T, for any linear transformation as [T] = [T(v1) | T(v2) | T(v3)], where v1, v2, v3 are standard basis vectors. Standard basis vectors are defined as the vectors of length 1 along the coordinate axes. The coordinate axes are always pairwise orthogonal. The column-vectors, T(v1), T(v2), T(v3), are transformed basis vectors. If we apply the linear transformation [T] to the standard basis vectors, they will remain pairwise orthogonal because each vector is rotated through the same angle around the same axis. Their lengths do not change because [T] is a simple rotation. So, the column-vectors of the standard matrix are pairwise orthogonal and their lengths are all 1. 2. Find the standard matrix of the linear transformation of R3 which reflects every vector V about the plane ax+by+cz=0 We know that the normal vector of this plane will have components identical to the coefficients from the equation for the plane. Thus, we can define a normal vector to the plane (along which we will want to project the head of the vector v) to be k(a,b,c) for some constant k. We apply this constant k to scale the vector in such a way as to have it be the vector from the plane to the head of vector v. Thus we have a triad of vectors: v (the original vector), w (the vector normal to the plane defined as k(a,b,c) ) and p (the vector that is the projection of v on the plane, and is v-w). We also know that vectors p and w are orthogonal to each other. We are trying to find the vector q = v-2w. Let us first, though, solve for k.

p=v-w p=(x-ka,y-kb,z-kc) p(a,b,c) = 0 (x-ka,y-kb,z-kc)(a,b,c) = 0 ax-ka2 + by-kb2 + cz - kc2 = 0 (ax + by + cz) - k(a2+b2+c2) = 0 (ax + by + cz) = k(a2+b2+c2) k = (ax + by + cz)/(a2+b2+c2) q = q = v-2k(a,b,c) (x-2ka, y-2kb, z-2kc)

put as a function, f(V) = (x-2a(ax + by + cz)/(a2+b2+c2), y-2b(ax + by + cz)/(a2+b2+c2), z-2c(ax + by + cz)/(a2+b2+c2))

The coefficients of x, y, z in this formula form the standard matrix. So the standard matrix is: > A:=matrix([[1-2*a^2/(a^2+b^2+c^2),-2*b*a/(a^2+b^2+c^2),-2*c*a/(a^2+b^2+c^2)],[2*a*b/(a^2+b^2+c^2),1-2*b^2/(a^2+b^2+c^2),-2*c*b/(a^2+b^2+c^2)],[2*a*c/(a^2+b^2+c^2),-2*b*c/(a^2+b^2+c^2),1-2*c^2/(a^2+b^2+c^2)]]); [ 1-2a2 / %1 -2ba / %1 -2ca / %1 ] A := [ -2ba / %1 1-2b2 / %1 -2cb / %1 ] [ -2ca / %1 -2cb / %1 1-2c2 / %1 ]
%1 := a2 + b2 + c 2

Check if the vector (1,2,3,4) belongs to the column space of the matrix [123] [456] [789] [012] A vector that belongs to the column space of this matrix is a linear combination of the three columns. So, we have the following system of equations:
x + 2y + 4x + 5y + 7x + 8y + y + 3z 6z 9z 2z = = = = 1 2 3 4.

This system has the following augmented matrix: > A:= matrix([[1,2,3,1],[4,5,6,2],[7,8,9,3],[0,1,2,4]]); [1231] [4562] A := [7893] [0124] > A:= gaussjord(A); [ 1 0 -1 0 ] [012 0] A := [000 1] [000 0] This vector does not belong to the matrix's vector space since the Gauss-Jordan form of the augmented matrix has no solution. Since column three is impossible. 0*x + 0*y + 0*z = 1 is impossible.

2. Find the null space of the matrix [12345] [23456] [10206] The kernel or null space of a matrix is the set of solutions of the homogeneous system of equations Av=0, where v = (x,y,z,t,s). The augmented matrix of this system is: [123450]

[234560] [102060] The Gauss-Jordan elimination procedure gives the following matrix: > gaussjord(B); [ 1 0 0 -4/3 0 0 ] [ 0 1 0 5/3 -2 0 ] [ 0 0 1 2/3 3 0 ] Hence, the general solution is
x y z t s = = = = = (4/3)*t (-5/3)*t + 2*s (-5/3)*t + 2*s t s,

or in vector notation: (x,y,z,t,s) = ((4/3)*t, (-5/3)*t + 2*s, (-5/3)*t + 2*s , t, s) = (4/3,-5/3,-5/3,1,0)t + (0,2,2,0,1)s.

So, the null space is spanned by the vectors (4/3,-5/3,-5/3,1,0) and (0,2,2,0,1). 3. Let p1 = x3 + x + 1, p2 = 2x3 - x2, p3 = x2 + x + 1, and q1 = x3 - 1, q2 = x3 - x2 + 1, q3 = x3 + x be polynomials. Is it true that span(p1,p2,p3) = span(q1,q2,q3) ? If span(p1,p2,p3) = span(q1,q2,q3) then p1, p2, and p3 can be represented as linear combinations of q1, q2, and q3, and vice versa. This is not true. It is not possible to form p1 from a linear combination of q1, q2, and q3. We can prove this by attempting to solve the following equation. p1 = a*q1 +b*q2 + c*q3 x3 + x + 1 = a(x3 - 1) + b(x3 - x2 + 1) + c(x3 + x) This can be written as the following system of equations:
a + b -b -a + b = = c = = 1 0 1 1

This is impossible. If b = 0, then equationa 1 and 4 have solutions for a, 1 and -1 respectively. So, span(p1,p2,p3) is not equal to span(q1,q2,q3). 4. Find the kernel of the linear transformation from R4 to R2 with the following standard matrix. To find the kernel of this matrix we use the same method as we used in problem #2. The augmented matrix of the system Av = 0 (v = (x,y,z,t)) is [12340] C := [21340] > gaussjord(C); [ 1 0 1 4/3 0 ] [ 0 1 1 4/3 0 ] So, the general solution is
x y z t = = = = -z - (4/3)*t -z - (4/3)*t z t,

or in vector notation: (x,y,z,t) = (-z-(4/3)*t, -z-(4/3)*t, z, t) = (-1,-1,1,0)z + (-4/3,-4/3,0,1)t.

The kernel is spanned by (-1,-1,1,0) and (-4/3,-4/3,0,1). 5. Is it true that the range of the transformation in the previous problem coincides with R2? The range of a transformation is spanned by its column vectors. So, it is spanned by the vectors (1,2), (2,1), (3,3), and (4,4). The vectors (1,0) and (0,1) span the R2. So, to prove that the range of the transformation in the previous problem coincides with R2, we need to prove that the span of the column vectors is equal to the span of the basic vectors, i and j. It is clear that the column vectors are spanned by the basic vectors i and j. Thus, it is only necessary to prove that i and j are spanned by the column vectors. We need to prove the following two systems of equations
x1* [1] [2] and x2* [1] [2] + y2* [2] [1] + z2* [3] [3] + t2* [4] = [4] [0] [1] + y1* [2] [1] + z1* [3] [3] + t1* [4] = [4] [1] [0]

have solutions. If they do, the column vectors span i and j. We can combine the augmented matrices of these two systems to get > A:= matrix([[1,2,3,4,1,0],[2,1,3,4,0,1]]); A := > gaussjord(A); [ 1 0 1 4/3 -1/3 2/3 ] [ 0 1 1 4/3 2/3 -1/3 ] Looking at the row-echelon form of the augmented matrix we can see that the two systems both have an infinite number of solutions. So, the basic vectors are indeed spanned by the column vectors and the range of the transformation in problem #4 coincides with R2. [123410] [213401]

Homework due Class 28 1. Check whether the following vectors in R5 are linearly independent. If they are not linearly independent, find one of them which is a linear combinationof the others. 1. (1,2, 3, 4, 5), (2,3,4,5,6), (1,0,0,0,1), (2, 5, 7, 9, 10); 2. (2,1,1,1,1), (2,1,4,4,4), (3,3,3,1,1). 2. Are these functions linearly independent as elements of C[0,1]: f(x)=x, g(x)=ex, h(x) = e2x. 1. Check whether the following vectors in R5 are linearly independent. If they are not linearly independent, find one of them which is a linear combination of the others. 1. (1,2, 3, 4, 5), (2,3,4,5,6), (1,0,0,0,1), (2, 5, 7, 9, 10); The set of vectors in R5 are linearly independent if the only solution to the equation below is the trivial (all zeros) solution: a(1,2,3,4,5)+b(2,3,4,5,6)+c(1,0,0,0,1)+d(2,5,7,9,10)=(0,0,0,0,0)

or
a+2b+c+2d 2a+3b+5d 3a+4b+7d 4a+5b+9d 5a+6b+c+10d = = = = = 0 0 0 0 0

>with(linalg); Warning: new definition for norm Warning: new definition for trace >A:=matrix([[1,2,1,2,0],[2,3,0,5,0],[3,4,0,7,0],[4,5,0,9,0],[5,6,1,10,0]]); [1212 0] [2305 0] A := [ 3 4 0 7 0 ] [4509 0] [ 5 6 1 10 0 ] >gaussjord(A); [1001 0] [0101 0] [ 0 0 1 -1 0 ] [0000 0] [0000 0]

The solutions of this system are:


a b c d = = = = -d -d d d

So the system has nontrivial solutions. Thus set of vectors in R5 form a linearly dependent set. Let us try to form (1,2,3,4,5) as linear combination of (2,3,4,5,6),(1,0,0,0,1) and (2,5,7,9,10). These vectors must then sastify the system of linear equation below: e(2,3,4,5,6)+f(1,0,0,0,1)+g(2,5,7,9,10)=(1,2,3,4,5)

or
2e+f+2g 3e+5g 4e+7g 5e+9g 6e+f+10g = = = = = 1 2 3 4 5

>B:=matrix([[2,1,2,1],[3,0,5,2],[4,0,7,3],[5,0,9,4],[6,1,10,5]]); [212 1] [305 2] B := [ 4 0 7 3 ] [509 4] [ 6 1 10 5 ] >gaussjord(B); [ 1 0 0 -1 ] [010 1] [001 1] [000 0] [000 0] the solutions of this system are:
e =- 1 f = 1 g = 1

So (1,2,3,4,5) is a linear combination of the others: (1,2,3,4,5) = -(2,3,4,5,6) + (1,0,0,0,1) + (2,5,7,9,10). 1. Check whether the following vectors in R5 are linearly independent. If they are not linearly independent, find one of them which is a linear combinationof the others. 2. (2,1,1,1,1), (2,1,4,4,4), (3,3,3,1,1). similarly, we construct of the system of equation: x1(2,1,1,1,1)+x2(2,1,4,4,4)+x3(3,3,3,1,1)=(0,0,0,0) or
2x1+2x2+3x3=0 x1+ x2+3x3=0 x1+4x2+3x3=0 x1+4x2+ x3=0 x1+4x2+ x3=0

>A:=matrix([[2,2,3,0],[1,1,3,0],[1,4,3,0],[1,4,1,0],[1,4,1,0]]); [2230] [1130] A := [ 1 4 3 0 ] [1410] [1410] >gaussjord(A); [1000] [0100] [0010] [0000] [0000] the solution of the system are:
x1=0 x2=0 x3=0

Thus the system has only one solution. Therefore, the vectors(2,1,1,1,1),(2,1,4,4,4) and (3,3,3,1,1)form a linearly independent set. 2. Are these functions linearly independent as elements of C[0,1]: f(x)=x, g(x)=ex, h(x) = e2x.

These functions are linearly independent if only the trivial solution sastifies the system below: k1f(x)+k2g(x)+k3h(x)=0 or k1x+k2*ex+k3*e2x=0 By substituting for specific value of x, we can generate a system of equations which could help show that these functions are linearly independent. At x=0: k2 + k3 = 0 At x=1: k1 + k2*e + k3*e2 = 0 At x=0.2: 0.2*k1 + k2*e0.2 + k3*e0.4 = 0 At x=0.3: 0.3*k1 + k2*e0.3 + k3*e0.6 = 0 Represented as a matrix: >A:=matrix([[0,1,1,0],[1,evalf(E^1),evalf(E^2),0],[0.2,evalf(E^0.2),evalf(E^0.4),0],[0.3,evalf (E^0.3),evalf(E^0.6),0]]); [0 1 1 0] [ 1 2.718281828 7.389056096 0 ] A := [ .2 1.221402758 1.491824698 0 ] [ .3 1.349858808 1.822118800 0 ] >gaussjord(A); [1000] [0100] [0010] [0000] As we can see, even for these limited values of x, the only solution is the trivial solution, so the equations are linearly independent.

1. Prove that if A1, ..., An are pairwise orthogonal vectors in a Euclidean vector space, then:
||A1 + A2 + ... + An||2 = ||A1||2 + ...+ ||An||2. (1)

By definition, in a Euclidean vector space, ||A||2 = A*A. So, equation (1) can be re-written as
(A1 + A2 + ... + An)*(A1 + A2 + ... + An) = A1*A1 + ... + An*An. (2)

The left side of this equation can be expanded, by the properties of the dot product in a Euclidean vector space. So, equation (2) becomes
A1*A1 + A1*A2 + ... + A2*A1 + A2*A2 + ... + An*An = A1*A1 + ... + An*An

But the vectors are pairwise orthogonal. Hence, every term that is not the dot product of a vector with itself will drop out on the left side (the dot product of two orthogonal vectors equals zero). The left side will become the right side. The proof is complete. 2. A function T from C[0,1] to R2 is given by the following formula: T(f(x)) = (f(0),f(1)).

Is it true that T is a linear transformation? To determine if T is a linear transformaiton, we must test to see if the two properties of a linear transformation hold for T. Property 1:
If f(x) and g(x) are two functions from C[0,1] then T(f(x) + g(x)) = ((f(0)+g(0), f(1)+g(1)) = (f(0),f(1)) + (g(0),g(1)) (by the rules of addition in R2) = T(f(x)) + T(g(x)).

So, the transformation T obeys the first property.

Property 2:
If f(x) is a function from C[0,1] and k is any number then

T(kf(x)) = (kf(0),kf(1)) R2) = k*(f(0),f(1)) (by the rule of scalar multiplication in = k*T(f(x)).

So, the transformation T also obeys the second property. Hence, T is a linear transformation. 3. Find the standard matrix of the linear operator in R3 which first reflects a vector about the plane 2x+3y+z=0 and then rotates it 60 degrees counterclockwise about the z-axis. In the homework we found the standard matrix of a reflection about the plane ax + by + cz = 0

to be [ (-a2 + b2 + c2)/%1 -2ab/%1 -2ac/%1 ] 2 2 2 A := [ -2ab/%1 (a - b + c )/%1 -2bc/%1 ] 2 2 [ -2ac/%1 -2bc/%1 (a + b - c2)/%1 ]
%1 := a2 + b2 + c2

If we plug in a=2, b=3, and c=1, we get the standard matrix [ 3/7 -6/7 -2/7 ] A := [ -6/7 -2/7 -3/7 ] [ -2/7 -3/7 6/7 ] The standard matrix for a counterclockwise rotation about the z-axis is similar to the rotation of an R2 vector in the xy-plane. We know during the rotation, the z-coordinate of the vector is left unchanged. Hence, the standard matrix is simply [ cos(a) -sin(a) 0 ] B := [ sin(a) cos(a) 0 ] [0 0 1] So, for a 60 degree rotation the standard matrix is [ 1/2 -1/2 * 31/2 0 ] B := [ 1/2 * 31/2 1/2 0] [0 0 1] Hence, the standard matrix of the linear operator that reflects a vector about the plane and then

rotates it 60 degrees is BA. > BA:= evalm(B&*A); [ 3/14 + 3/7 * 31/2 -3/7 + 1/7 * 31/2 -1/7 + 3/14 * 31/2 ] BA := [ 3/14 * 31/2 - 3/7 -3/7 * 31/2 - 1/7 -1/7 * 31/2 - 3/14 ] [ -2/7 -3/7 6/7 ] 4. Find the vectors which span the null space of the following matrix: [23456] A := [12354] The null space of this matrix consists of all vectors, v = (x,y,z,t,s), which satisfies the equation Av=0. This is a system of equations with standard matrix [234560] B := [123540] with reduced row-echelon form [ 1 0 -1 -5 0 0 ] [012 5 20] The general solution of the system is
x y z t s = = = = = z + 5t -2z - 5t - 2s z t s,

or in vector notation (x,y,z,t,s) = (z+5t, -2z-5t-2s, z, t, s) = z(1,-2,1,0,0) + t(5,-5,0,1,0) + s(0,-2,0,0,1).

So, the null space is spanned by the vectors (1,-2,1,0,0), (5,-5,0,1,0), and (0,-2,0,0,1). 5. What are the standard matrix and the range of the transformation T from R4 to R2 which takes every vector (x,y,z,t) to (2x+3y,x+4y+z+t)? The transformation goes from R4 to R2 so the standard matrix must be a 2 by 4 matrix. It is simply the matrix with the coefficients of the terms in the result of the transformation as rows. Hence, the standard matrix, T, is [2300] T := [1411]

> v:= matrix([[x],[y],[z],[t]]); [x] [y] v := [z] [t] > b:=evalm(T&*v);


b := [ [ 2x + 3y ] x + 4y + z + t ]

The range consists of all vectors, b, in R2, such that there is a vector v, in R4, that is a solution to the equation Tv = b. It is spanned by the column vectors of T. So, the range of the transformation is the set of all linear combinations of the column vectors of T. We can guess that the range is all of R2. This would mean that the span of the column vectors is equal to the span of the basic vectors i and j in R2. By definition, if the spans of these two subsets are equal then the vectors of one subset are linear combinations of vectors from the other subset and vice-versa. It is obvious that the column vectors of T are linear combinations of i and j, so, all that is left is to prove that i and j are linear combinations of the column vectors of T. We have two systems of equations: a* and a* [2] [3] [0] [0] [0] +b* +c* +d* = [1] [4] [1] [1] [1] [2] [3] [0] [0] [1] +b* +c* +d* = [1] [4] [1] [1] [0]

These systems have the augmented matrix F := with reduced row-echelon form [ 1 0 -3/5 -3/5 4/5 -3/5 ] [ 0 1 2/5 2/5 -1/5 2/5 ] It can be seen, from the reduced row-echelon form of the augmented matrice for the two systems, that both of these systems have solutions. In fact, they both have an infinite number of solutions. So, the span of the column vectors is equal to the span of the basic vectors i and j. Hence, the range of the transformation T is all of R2. [230010] [141101]

Вам также может понравиться