Re: 'help matrix' -- Re: [Bug #1264105] In currrent CVS version plotting matrix is broken

[Ethan A M. <merritt@u.washington.edu>]

On Sunday 23 October 2005 10:38 pm, mi...@ph... wrote:
> > "Column 3" may or may not be a Z value, depending
> > on what kind of graph you are describing.
>
> I would keep the x, y and z for both plot and splot cases, and add that
> the default using for plot is
>     plot 'a.dat' matrix using 1:3
> and not using 1:2 which makes no good sense for matrices.

Sorry, I must be really dense today.
I am just not understanding your way of naming things.

In a 2D plot we are plotting y = f(x).
x is taken from the first using spec.
y is taken from the second using spec.
There is no z.

So when you say   "plot 'a.dat' matrix using 1:3"
you are asking for x from column 1 and y from column 3.
How can it be correct to document column 3 as being z
in this case?


> >   Matrix values are read in a row at a time, i. e.,
> >           M11 M12 M13 M14 ...   (i=1)
> >           M21 M22 M23 M24 ...   (i=2)
> >           M31 M32 M33 M34 ...   (i=3)
> >           ... ... ... ... Mij
> >
> >   The interpretation of columns in the "using" specification for
> >   matrix data is different from the usual case. "$1" refers to the
> >   matrix row index i, "$2" refers to the matrix column index j,
> >   and "$3" refers to the matrix entry Mij.
>
> Indexing starts from 0. 

Oops. Yes.

> Further, x is the column index j, y is row index i.

Yeah, I'm not doing so well in describing this either.
And on top of that we may be suffering from a difference in
naming conventions.  I am now uncertain whether the term
"row index" means the same thing to you that it does to me.
And even if it does, there are people in the world that use
a different convention so maybe it is better not to use the
terms at all.

I agree that the x value of the plot is taken from the index j.

But neither i nor j is "y" in the sense of a 2D graph coordinate.
The y value is take from Mij.

If there are multiple lines in the data file, then i will be
counting out successive plots in the 2D graph. So it is neither
"x" nor "y", but rather it is "plot number".

Maybe we need to write totally separate sections for the 2D case
and the 3D case.

Anyhow, I don't seem to be having much luck at writing this
section. I'll hand it back to you.
Or maybe somebody else wants to take a stab at it?

-- 
Ethan A Merritt
Biomolecular Structure Center
University of Washington, Seattle 98195-7742