Re: Graphing timetables in excel

[G.Lambert@unsw.edu.au (Geoff Lambert)]

David Johnson <trainman@ozemail.com.au> wrote:

>Geoff Lambert wrote:

>> My program also does a number of statistical analyses on the traffic-
>> number of crosses, percentage line occupancy, etc.  Also the details
>> of which train meets which on single lines.  The program accepts an
>> entire timetable (e.g. a week's worth) and allows you to construct a
>> graph for any day of the week.  It can handle conditional trains.
>> Train numbers, etc are printed as text close to their relevant lines.

>Are the stations referenced as kilometrage from Sydney, like a train
>controller's graph?  If so, how does your program handle one track being longer
>than the other (Bethungra, Frampton)?

In such a case, the line representing a train on the longer track,
takes off in a 3rd (and sometimes a 4th) dimension- effectively it
bores its way down through the paper, orbits a black hole for a period
of time and then re-emerges onto the paper.  People on the trains that
follow this line age slightly less than those on trains that stay on
the paper, due to the Fitzgerald contraction.  In the case of spiral
lines like Bethungra extra relativistic effects occur.

Seriously, though, this is not easy to do- a method such as that used
to expand station-yards in some train graphs could be used.  At any
rate, there is no a priori need to show them spatially realistically
or with a distance scale that is perfectly regular- the first because
this would ipso facto be a double line- no crosses (or collisions!)
can occur and  in the second case, a graph of any scale, even a
variable one, preserves the topological relationships of the graph.
It's just that most people like to be able to judge the speed of a
train by the slope of the line.  But even in practical train graphs,
this is distorted for convenience sake (in stations, as mentioned B4)

Geoff Lambert