The Student Room Group

Complex numbers

Not sure how to translate the point (2,3) back to the originScreenshot 2022-05-11 at 16.12.00.png
(edited 1 year ago)
Original post by Student 999
Not sure how to translate the point (2,3) back to the originScreenshot 2022-05-11 at 16.12.00.png

What have you done so far? Have done the cube roots of "unity" then rotation and translation? Or ...
(edited 1 year ago)
Original post by mqb2766
What have you done so far? Have done the cube roots of "unity" then rotation and translation? Or ...

I've worked out the roots of unity for triangle centre about origin, but can't seem to figure how to translate the origin to (2,3) so that I can apply it to the other roots
Original post by Student 999
I've worked out the roots of unity for triangle centre about origin, but can't seem to figure how to translate the origin to (2,3) so that I can apply it to the other roots

If z3=1z^3 = 1 provides the cubes of unity about the origin, then (zz0)3=1(z-z_0)^3 = 1 provides the cubes of unity about the point z0z_0. Set it appropriately to get the centre at (2,3) on the complex plane.

But the easier approach would just be to see what is the complex number (i.e. vector) that takes you from (2,3) to (3,-2) and then just rotate it 120 degrees by multiplication by an appropriate complex number. Then add the rotated vector onto the center in order to go straight to where another vertex is.
(edited 1 year ago)
Original post by Student 999
I've worked out the roots of unity for triangle centre about origin, but can't seem to figure how to translate the origin to (2,3) so that I can apply it to the other roots


Add (2,3) to each root. Im assuming that youve scaled by 26^1/2 and rotated by the correct arg first.

Note once you know the arg rotation, you could almost write down the correct trig expressions without doing any complex stuff.
(edited 1 year ago)
Original post by RDKGames
If z3=1z^3 = 1 provides the cubes of unity about the origin, then (zz0)3=1(z-z_0)^3 = 1 provides the cubes of unity about the point z0z_0. Set it appropriately to get the centre at (2,3) on the complex plane.

But the easier approach would just be to see what is the complex number (i.e. vector) that takes you from (2,3) to (3,-2) and then just rotate it 120 degrees by multiplication by an appropriate complex number. Then add the rotated vector onto the center in order to go straight to where another vertex is.


Not quite sure how your first method works/derived however following your second method I was able to get the correct answer
Original post by mqb2766
Add (2,3) to each root. Im assuming that youve scaled by 26^1/2 and rotated by the correct arg first.

Note once you know the arg rotation, you could almost write down the correct trig expressions without doing any complex stuff.

Having drawn a simple sketch I realised that the transformations compromises as enlargement and rotation then addition, however I'm a bit unsure as to how you would find the rotation, Is it essentially just finding the angle difference between e^(0i) and the vertex given in the question then using modulus argument form you can correct the orientation then enlarge before adding the complex number 2+3i
Original post by Student 999
Having drawn a simple sketch I realised that the transformations compromises as enlargement and rotation then addition, however I'm a bit unsure as to how you would find the rotation, Is it essentially just finding the angle difference between e^(0i) and the vertex given in the question then using modulus argument form you can correct the orientation then enlarge before adding the complex number 2+3i


You know one vertex, relative to the center is (1,-5) which is arg atan(-5). So if you rotate the x-axis "root" by that amount, then all three "roots" would be rotated by that. Or alternatively add 2pi/3 and 4pi/3 to that angle and forget about the complex dressing and just use simple trig.
(edited 1 year ago)
Original post by mqb2766
You know one vertex, relative to the center is (1,-5) which is arg atan(-5). So if you rotate the x-axis "root" by that amount, then all three "roots" would be rotated by that. Or alternatively add 2pi/3 and 4pi/3 to that angle and forget about the complex dressing and just use simple trig.

IMG_128F2B241604-1.jpeg This what I've done which is what I think you mean
Original post by Student 999
IMG_128F2B241604-1.jpeg This what I've done which is what I think you mean

a = atan(-5), r = sqrt(26), c=(2,3)
z1 = c + r*e^(ai)
z2 = c + r*e^((a+2pi/3)i)
z3 = c + r*e^((a+4pi/3)i)
Tbh, there is little complex numbers here, Im just using it to avoid writing cis (cos(),sin())
Original post by mqb2766
a = atan(-5), r = sqrt(26), c=(2,3)
z1 = c + r*e^(ai)
z2 = c + r*e^((a+2pi/3)i)
z3 = c + r*e^((a+4pi/3)i)
Tbh, there is little complex numbers here, Im just using it to avoid writing cis (cos(),sin())

IMG_1283.jpgScreenshot 2022-05-11 at 19.23.28.png Is this the correct approach, are there any other methods?
Original post by Student 999
IMG_1283.jpgScreenshot 2022-05-11 at 19.23.28.png Is this the correct approach, are there any other methods?

Looks good. When you do the e^i -> sin transformation, Id probably have factorized e^i4pi/9 on the numerator and e^i pi/9 on the denominator, so centered each to get the appropriate sin(), then divided those two imaginary exponentials to get your e^i pi/3. But thats minor. Also you could have written the sum as
(e^i8pi/9 - 1) / (e^i2pi/9 - 1)
as that leaves the usual sin() represntation when you factorize the two imaginary exponentials out of the numerator and denominator.

You could have done it geometrically as well, by forming triangles with the center of the (part) polygon. Its arguably a bit easier to do this.
(edited 1 year ago)

Quick Reply

Latest