Hi Geoff,

This post comes by way of Matthew Graham <[log in to unmask]>, a 
regular contributer to FISH Technical.

-------
Subject: Plotting spherical coordinates
From: Matthew Graham <[log in to unmask]>
Date: Wed, 04 May 2005 16:25:47 -0700
To: [log in to unmask]

Hi,

A paper that I find handy for all spherical map projections used in 
astronomy is:

"Representations of celestial coordinates in FITS" by M. R. Calabretta 
and E. W. Greisen, Astronomy and Astrophysics, 395, 1077-1122 (2002)

It gives the maths for the transformations and also shows the 
projections graphically. Recording your 3D data would then just be a 
matter of plotting a point on whatever projection you are using.

A very common problem in astronomy is also determining whether the 
distribution of objects you are considering is homogeneous and isotropic 
or whether there are inhomogeneities or anisotropies and what their 
scale is. A textbook I would recommend for reference is:

"Statistics of the Galaxy Distribution" by Vicent Martinez and Enn Saar

    Cheers,

    Matthew


Geoff Carver wrote:
> the usual apologies for cross-posting
> i'm doing some background preparation for some work that i hope will eventually allow me to model/simulate post-depositional transformations of the (archaeological) burial assemblage (and yes, this will definitely include bayesian stats)
> i'm looking for ways that have been tried for recording directions in 3D (a spherical equivalent of the direction-rose sort of plots you get for recording vectors; things like wind direction & so on: so many days with the wind between 0 and 10°, so many between 10 and 20°, etc.)
> does anyone have examples of:
> first: this kind of recording
> and second: ways for doing statistical analysis of direction in 3D space?
> among other things i'm wondering if/how you can get around the 0/360° problem...