[yt-users] Profiles and normal graphs

[Matthew Turk]

Hi,

I'd like to echo a bit of Brian and Britton's response.  In YT a
'profile' can be in multiple dimensions; it's a way of taking an
N-dimensional weighted-average of the data, after having segmented it
into different bins.

For instance, in my work, I often need to examine how the molecular
hydrogen fraction in gas is related to the density and temperature.
So I take a weighted-average -- what in YT is called a 'phase profile'
-- of the gas, looking at the average H2 fraction in bins of Density
and Temperature.

However, that's a means of reducing the dimensionality of the data.
If you want to examine spatial information, you would usually use
something like a slice (whereby you examine all of the data in an
axis-parallel fashion at a given point along the axis) or a projection
(where the data is summed along a given axis, or a weighted-average
along a given axis is taken.)

Again, though, if you tell us a bit about the question you're trying
to answer, we can possibly respond with more meaningful information
about ways of approaching that question.

Best wishes,

Matt

On Thu, Oct 2, 2008 at 5:51 PM, Brian O'Shea <bwoshea at gmail.com> wrote:
> The term profile is usually short for radial profile.  Making a radial
> profile starts by choosing some point in the simulation box, then
> placing the grid cells into groups based on their distance away from
> that point.  In essence, each group of cells represents a spherical
> shell of some thickness.  Once we have the cells in groups, then we
> take the average of a certain quantity, say density or temperature, in
> each of those groups.  Then, what we have is simply the average value
> of some quantity as a function of distance away from the central
> point.  Examples of radial profiles made from Enzo simulation data can
> be seen in this paper:
> http://adsabs.harvard.edu/abs/2007ApJ...654...66O (figure 4, for
> example).
>
> A graph, on the other hand, isn't really anything in particular.
> There are many ways to visualize data from simulations, and making
> radial profiles is just one of them.  I encourage you to check out the
> YT analysis toolkit (yt.enzotools.org) for various other ways of
> looking at your data.  Different problems are better visualized in
> different ways.  Perhaps if you explain what type of problem you're
> working on, we can give you some direction.
>
> Cheers,
>
> Britton
>
> On Thu, Oct 2, 2008 at 8:46 PM, rsoares <dlleuz at xmission.com> wrote:
>> Hi,
>>
>> What's the difference b/t a "density profile" or  "temperature
>> profile" etc and just a basic density graph, temperature graph, etc?
>>
>> (Cosmological simulations).
>>
>> Is a profile a graph over a chunk of space which is 2-dimensional plus
>> width-not-too-small /large, whereas a graph is just a 2-dimensional
>> plot, y vs. x, say.
>>
>> Is this chunk what is called a slab, in numerical simulations?
>>
>> And how does a spherically-average profile differ from a chunk or slab
>> profile?
>>
>>
>> Thanks.
>> -R.Soares
>> _______________________________________________
>> yt-users mailing list
>> yt-users at lists.spacepope.org
>> http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
>>
> _______________________________________________
> yt-users mailing list
> yt-users at lists.spacepope.org
> http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
>