1
$\begingroup$

Let $R_\theta(v)$ be the result of rotating a point $v = (x, y)$ around the origin through an angle of $\theta$.

Then $R_\theta(v) = (x\cos\theta-y\sin\theta, x\sin\theta + y\cos\theta)$.

Now given diagram like this:

enter image description here

(From Calculus/Spivak)

Show that:

$$x'=\frac{1}{\sqrt{2}}x+\frac{1}{\sqrt{2}}y$$ $$y'=-\frac{1}{\sqrt{2}}x+\frac{1}{\sqrt{2}}y$$

Intuitively that makes sense, but I haven't been able to express the sides $x',y'$ using some expression.

If we start with some point $v =(x,y)$ and rotate it by $45^\circ$, then:

$$R_\theta(v) = (x\cos\theta-y\sin\theta, x\sin\theta + y\cos\theta) \\ = (x\frac{1}{\sqrt{2}}-y\frac{1}{\sqrt{2}}, x\frac{1}{\sqrt{2}} + y\frac{1}{\sqrt{2}})$$

And if we refer to $x',y'$ as vectors, then:

$$v = x' + y'$$

But I can't bring my thoughts together. How can I express the sides of the rectangle in terms of $x, y$?

$\endgroup$
3
  • $\begingroup$ rotate the extreme points of the side, the the segment with extreme the images will be the new side $\endgroup$
    – ALG
    Oct 30, 2018 at 11:46
  • $\begingroup$ maybe try looking at this from a matrix point of view? That's what came to my mind first. $\endgroup$
    – Landuros
    Oct 30, 2018 at 11:54
  • $\begingroup$ Forgot to mention, the book I follow didn’t introduce matrices yet, but vectors. So I’m looking for vector/trig answers. $\endgroup$
    – BMBM
    Oct 30, 2018 at 23:29

1 Answer 1

1
$\begingroup$

One thing is to keep the point fixed and rotate the the axes, another is to rotate the point keeping the axes fixed.

Clearly you pass from one to the other by inverting the sign of the angle.

The matrix $$ {\bf R}_{\,\,z} (\alpha) = \left( {\matrix{ {\cos \alpha } & { - \sin \alpha } \cr {\sin \alpha } & {\cos \alpha } \cr } } \right) $$ when applied to a vertical vector (point or axis), returns the vector rotated by $\alpha$ in the old (i.e. fixed) coordinates.

While $$ {\bf R}_{\,\,z} ( - \alpha) = {\bf R}_{\,\,z} ^{\,T} (\alpha) = {\bf R}_{\,\,z} ^{\, - \,{\bf 1}} (\alpha) $$ when applied to a vertical vector gives that vector (fixed) in the new coordinates rotated by $+\alpha$
(or the vector rotated by $-\alpha$, keeping reference fixed).

And that is what you were asked to prove.

Told in other terms, without using the concept of matrix, if we indicate by $\bf i,\, \bf j$ and $\bf {i'},\, \bf {j'}$ the unit reference vectors, and if (as it is usually done) you start with determining the new vectors referenced to the old, which upon a rotation of $+\alpha$ gives the system on the LHS $$ \left\{ \matrix{ {\bf i'} = \cos \alpha \,{\bf i} + \sin \alpha {\kern 1pt} {\bf j} \hfill \cr {\bf j'} = - \sin \alpha \,{\bf i} + \cos \alpha {\kern 1pt} {\bf j} \hfill \cr} \right.\quad \Leftrightarrow \quad \left\{ \matrix{ {\bf i} = \cos \alpha \,{\bf i'} - \sin \alpha {\kern 1pt} {\bf j'} \hfill \cr {\kern 1pt} {\bf j} = \sin \alpha \,{\bf i'} + \cos \alpha {\kern 1pt} {\bf j'} \hfill \cr} \right. $$ and which, when solved for $\bf i,\, \bf j$ provides that on the RHS.

Now the same vector $\bf v$ ( a "fixed" vector) will be expressed in the two references as $$ {\bf v} = x\,{\bf i} + y{\kern 1pt} {\bf j} = x'{\bf i'} + y'{\bf j'} $$ Then $$ \eqalign{ & {\bf v} = x'{\bf i'} + y'{\bf j'} = x\,{\bf i} + y{\kern 1pt} {\bf j} = \cr & = x\left( {\cos \alpha \,{\bf i'} - \sin \alpha {\kern 1pt} {\bf j'}} \right) + y\left( {\sin \alpha \,{\bf i'} + \cos \alpha {\kern 1pt} {\bf j'}} \right)\quad \Rightarrow \cr & \Rightarrow \quad \left\{ \matrix{ x' = x\cos \alpha + y\sin \alpha \hfill \cr y' = - x\sin \alpha + y\cos \alpha \hfill \cr} \right. \cr} $$

That is: $$ \left( {\matrix{ {x'} \cr {y'} \cr } } \right) = \left( {\matrix{ {\cos \alpha } & {\sin \alpha } \cr { - \sin \alpha } & {\cos \alpha } \cr } } \right)\left( {\matrix{ x \cr y \cr } } \right) = {\bf R}_{\,\,z} ( - \alpha)\left( {\matrix{ x \cr y \cr } } \right) $$ as told above (in matrix terms).

$\endgroup$
4
  • $\begingroup$ Sorry, I forgot to mention, I wasn't formally introduced to matrices yet. Do you have a non-matrix explanation? $\endgroup$
    – BMBM
    Oct 31, 2018 at 2:06
  • $\begingroup$ @Max no problem, I added a "vectorial" version, and so you are getting a brief introduction to matrices as well. $\endgroup$
    – G Cab
    Oct 31, 2018 at 14:30
  • $\begingroup$ Thank you so much! I understand now how the proof concludes. The only thing I’m missing still is how you create the connection between the new and old vectors. Like at the part where you say: ... when solved for i, j provides that on the RHS. It’s just algebraic manipulation of the LHS to get the RHS, or how can you obtain the two equations for i, j on the right hand side? $\endgroup$
    – BMBM
    Nov 1, 2018 at 0:41
  • 1
    $\begingroup$ @Max: it is just the solution of a linear system of two equations in two unknowns (i,j), given (i',j') as known (or as parameters). $\endgroup$
    – G Cab
    Nov 1, 2018 at 1:23

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .