Vector projection of u onto v (introduction & example)

Поділитися
Вставка
  • Опубліковано 5 лют 2025

КОМЕНТАРІ • 25

  • @Willigfhsty
    @Willigfhsty 3 місяці тому +3

    Took me way too long to find a decent video on how the projection vector formula is derived. Thank you very much for explaining it in such a clear and concise manner!

  • @jeffreyhunter4115
    @jeffreyhunter4115 2 місяці тому +2

    This is one of the best videos on vector projection I’ve watched. Excellent introduction with step-by-step explanations. Thank you for sharing!

  • @stigastondogg730
    @stigastondogg730 6 місяців тому +2

    Have used this to great affect with a digital compass I’m making at home.
    For those who aren’t aware, earths magnetic field is actually 3 dimensional. The sensor may not be aligned perfectly to the “world coordinates”……but I can get the vector of magnetic field measurement, then dot product against a vertical unit vector and remove from the signal, so I’m left with horizontal measurements.
    For product against east and north unit vectors, gives me magnitudes in each of those directions.
    Then I can calculate magnetic north direction and display as an angle on the little LCD screen.
    This stuff does have practical applications in the real world. Love it

    • @stigastondogg730
      @stigastondogg730 6 місяців тому +1

      And before someone says about cross product (lol) I do realise that. And have used it too…..but was conveying this part of my story as it relates to this video

  • @ianfowler9340
    @ianfowler9340 6 місяців тому

    Loved your last couple of videos on dot product. When I taught this I always started with the definition as : (1) u.v = |u| |v| cos(t) rather than the (2) u1v1 + u2v2 which I saw a a consequence of the definition. I know that (2) follows the rules of definitions but here is why I chose (1) as a more basic logical starting place. (for me anyway)
    There are several concepts in physics (which my senior students were already familiar with) that require multiplying 2 vector quantities to arrive at a meaningful scalar. The most common example in high school physics is that the work done by an unbalanced force is given by
    (3) W = |the component of F in the direction of the displacement| * |dispalcement|.
    Assuming F and D have the same initial point and separated by an angle t, you need to drop a perpendicular from the tip of F to vector D. You now have 2 vertical componenets of F - one in the direction of D and the other perpendicular to D (which does no work on the moving mass).This turns (3) into W = |F|cos(t)|*|D| = |F| |D| cos(t) which then gets defined as "dot product of F and D". Notice that you can drop the perpendicular from the tip of D to meet F and proceed in exactly the same way. If t is acute we get a positive work which then increases the speed (KE) in the direction of D. If t is obtuse we think of negative work as slowing the speed or increasing it in the opposite direction to D.
    I then did exactly what you did with the cosine law to arrive at the componenet form, u1v1 + u2v2, of the dot product imposed on a co-orinate system. This form, as you pointed out, gives us a more convenient/faster way of finding the angle between 2 vectors.
    Now that we have the "scalar projection of u on v", (magnitude) we now multiply that by a unit vector in the direction of v, and out pops the "vector projection of u on v" formula as v.v = |v|^2
    I really like your approach and as always - keep up the good work. So much fun.

  • @ejsafara456
    @ejsafara456 Місяць тому

    oh my god, such a good video, thank you!

  • @ianfowler9340
    @ianfowler9340 6 місяців тому

    I also used to verify the following with my students. Not needed but nice to see it actually works!: u = , v = , p =
    Your dotted line vector is u - p =
    so that p "dot" (u - p) = -232/81 + 152/81 + 80/81 = 0 Verifying that p and (u-p) are indeed perpendicular.

  • @isaaclearningtominecraft4751
    @isaaclearningtominecraft4751 6 місяців тому +4

    As somebody who loves geometric algebra, I've to say this: all you need is to notice (proj u)v=(proj u).v=u.v, so proj u=u.v/v.

    • @appybane8481
      @appybane8481 6 місяців тому

      It's ok to use u•v/v here .However, In general multivectors don't commute, so you should write (u•v)(v**-1)

    • @isaacto8761
      @isaacto8761 6 місяців тому

      ​@@appybane8481 Thanks. I just agree with the book of Alan Macdonald to always mean post-multiply with the inverse when using the division notation.

  • @phill3986
    @phill3986 6 місяців тому +6

    Umm not sure about the magnitude part at the end. Wouldn't the 4/9 need to be multiplied by the magnitude of v?

    • @chitlitlah
      @chitlitlah 6 місяців тому

      Yeah, it doesn't seem right to me either. Without doing any calculations, the magnitude has to be at least the absolute value of the largest coordinate, which is 8/9 in this case.

    • @simonecappellaro7075
      @simonecappellaro7075 6 місяців тому

      It’s wrong in fact. The magnitude of the projection is u*v divided by the magnitude of v. It’s 4/3

    • @BrettPienaar
      @BrettPienaar 6 місяців тому

      The magnitude is actually 12/9

    • @BrettPienaar
      @BrettPienaar 6 місяців тому

      @simoncaprello...correct

    • @bprpcalculusbasics
      @bprpcalculusbasics  6 місяців тому

      Thank you for pointing that out! I just trimmed out that part of the video!

  • @lesliesusil4711
    @lesliesusil4711 6 місяців тому +1

    Kindly explain me why the inventor mathematician defined a dot product as a scalar where as the cross product as a vector which is perpendicular to the both vectors and it's detection is depend on the right hand finger rule.

    • @ianfowler9340
      @ianfowler9340 6 місяців тому

      See my long comment above. Your answer might be there.

  • @gitboyyy
    @gitboyyy 6 місяців тому +2

    dude , can you do a video on inner product next and what exactly does it do ? i've heard it's similar to dot product and it kinda makes sense with summation being similar to integration but I just don't get it at all . thanks ☺

    • @phill3986
      @phill3986 6 місяців тому +1

      As I remember, dot product is a particular kind of inner product. Inner products can be done on functions

    • @ianfowler9340
      @ianfowler9340 6 місяців тому

      For 2 vectors -->u and -->v, the dot product calculates: (magnitude of the componenet of of -->u in the direction of -->v) multiplied by the (magnitude of -->v) .
      That is, | -->u|*|-->v|*cos(t). Very important scalar in physics. For example: W (scalar) = -->F "dot" -->d .

  • @urluberlu2757
    @urluberlu2757 6 місяців тому

    👍

  • @experimentbysaifanali7580
    @experimentbysaifanali7580 6 місяців тому

    First!