[gmx-users] manual suggestion from Argyris Karantrantos

[Mark Abraham]

Hi Argyrios,

In order that people can follow threads more easily, please write a new
email to the gromacs list, rather than reply to an existing one. People
may choose not to read your email because they're not interested in that
thread, and then you confuse both discussions by having them look the
same.

> Hello, Gromacs administrators,
>  I have a suggestion. In the online manuals can you explain how to read
> the output or log file after explaining how to use each command. For
> example,
>
>  g_hbond
>  Description
>  Flags
>  I/O
>  How to read outputs *

I think that would be overkill, because (with the possible exception of
.xpm files that I know nothing about) all of the gromacs files you'd
probably think of as "output" are text files, and gmxdump renders the
binary .tpr, .trr, .edr and .xtc into text form readily. There's nothing
to lose by trying "less file.ext" and seeing what's there - mostly it will
prompt you if it thinks the file is binary and give you the option of not
looking at gibberish.

It's also a bit anti-philosophy for Unix... there's almost never "one true
way" to do something. Users are expected to solve their problems in a
flexible manner, and "suggesting" how to read the outputs could restrict
the way people think about how to deal with the files they have generated.

> By the way, how do I read a .xvg from a G_hbond calculation?

Well, you can read it with anything really - it's just a text file
containing some comments, some header lines and then columns of
space-delimited numbers. It's designed to be read by the plotting program
Grace (as touched on in the opening paragraph to chapter 8), but it's not
hard to massage it for consumption by gnuplot, or even to import it into
horrible things like Excel :-O

Mark