Question: invalid calculation of Point3D transformation? about euclid HOT 4 CLOSED

cc @nical

from euclid.

Hi @youngec, if you look at transform_point3d, euclid treats points as if they had their homogeneous coordinate equal to one, and then in vector_3d a few lines below the computation corresponds to an homogeneous coordinate of zero, just like you said.

The issue is coming from the matrix in your test case: the translation components should be the members on the last row of the matrix and not the right column. You can try the same test again using the column_major notation:

    let m: Transform3D<f64> = Transform3D::column_major(
        1.0, 0.0, 0.0, 0.0,
        0.0, 1.0, 0.0, 0.0,
        0.0, 0.0, 1.0, -5.0,
        0.0, 0.0, 0.0, 1.0
    );

Or better even, use:

 let m: Transform3D<f64> = Transform3D::create_translation(0.0, 0.0, -5.0);

...to avoid the classical confusion around transposition.

from euclid.

Yikes, this always happens to me; getting row- and column-major orders confused. Unfortunately, I have to rely on specifying all components of the matrix by hand, so your second suggestion will not work in my case. Thanks for clearing that up! You can close the issue.

from euclid.

Thanks for checking!

from euclid.