#304 error in the documentation, version 3.0.0 (probably older)

[Jürgen Koslowski]

In Section 25.3 on page 361 the effect of the coordinate transformation cm is described as multiplying a matrix with a column vector from the right, and adding an offset in form of another column vector. In fact, the example following after the definition of cm only makes sense, when row vectors are used (with the same matrix).

[percusse]

It is easier to state that the matrix should be transposed in the documentation. So the affine map is $y = (A^T)*x + d$ where A is the documented matrix x is the old coord vector d is the translation and y is the transformed coordinate vector.

[Jürgen Koslowski]

"easier" maybe in the sense of making the correction (add a
transposition symbol to the matrix), but the pgf syntax uses
row-vectors rather than column-vectors, so in my view transposing
everything, vectors and the matrix, is bound to create confusion. But
the main thing, of course, is to have a correct description.

-- Jürgen

Quoting percusse talopnahli@users.sf.net:

It is easier to state that the matrix should be transposed in the
documentation. So the affine map is $y = (A^T)*x + d$ where A is the
documented matrix x is the old coord vector d is the translation and
y is the transformed coordinate vector.

---

[bugs:#304] error in the documentation, version 3.0.0 (probably older)

Status: open
Group: v1.0 (example)
Labels: documentation cm
Created: Sat Mar 22, 2014 10:17 PM UTC by Jürgen Koslowski
Last Updated: Sat Mar 22, 2014 10:17 PM UTC
Owner: nobody

In Section 25.3 on page 361 the effect of the coordinate
transformation cm is described as multiplying a matrix with a column
vector from the right, and adding an offset in form of another
column vector. In fact, the example following after the definition
of cm only makes sense, when row vectors are used (with the same
matrix).

---

Sent from sourceforge.net because you indicated interest in
https://sourceforge.net/p/pgf/bugs/304/

To unsubscribe from further messages, please visit
https://sourceforge.net/auth/subscriptions/

Juergen Koslowski If I don't see you no more on this world
ITI, TU Braunschweig I'll meet you on the next one
koslowj@iti.cs.tu-bs.de and don't be late!
http://www.iti.cs.tu-bs.de/~koslowj Jimi Hendrix (Voodoo Child, SR)

---

This message was sent using IMP, the Internet Messaging Program.

[percusse]

True but I don't see any row-vector convention in the manual other than writing the coordinates inline in Sec. 103.2.1. In fact PostScript uses col vectors. Hence my suggestion to fix the matrix.

[Jürgen Koslowski]

Quoting percusse talopnahli@users.sf.net:

True but I don't see any row-vector convention in the manual other
than writing the coordinates inline in Sec. 103.2.1. In fact
PostScript uses col vectors. Hence my suggestion to fix the matrix.

Right, in order to stay consistent with existing notation, that is
probably the way to go. Tells you something about the existing
notation, though... PostScript should have known better, its xf, not
f(x)! But I digress.

-- Jürgen

--
Juergen Koslowski If I don't see you no more on this world
ITI, TU Braunschweig I'll meet you on the next one
koslowj@iti.cs.tu-bs.de and don't be late!
http://www.iti.cs.tu-bs.de/~koslowj Jimi Hendrix (Voodoo Child, SR)

---

This message was sent using IMP, the Internet Messaging Program.

[Kpym]

This looks the same as the bug #339 (that is closed now).

[Christian Feuersänger]

Yes, it is a duplicate. Thanks for pointing this out.