[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [igraph] how to plot a very basic MST with igraph
From: |
Steve Boudreault |
Subject: |
Re: [igraph] how to plot a very basic MST with igraph |
Date: |
Wed, 5 Mar 2014 17:17:31 +0100 |
User-agent: |
SquirrelMail/1.4.21 |
>
> But I'd prefer you tell us want you want to do, and then we don't need
> to
> guess.
>
Please look at the following image :
http://inspirehep.net/record/833443/files/mst.png
These are "minimum spanning tree" (right?) of stars distributed in X (i.e.
right ascension, RA) and Y (i.e. declination, DEC). You have this for
three open clusters : IC2391, M34, M11.
This *is* what I want to do. I want to do a plot like this.
I have a list of objects in RA and DEC :
alpa (RA) delta (DEC)
0.923 -12.345
0.912 -11.222
0.812 -11.763
0.902 -11.923
0.942 -12.001
... ...
I would like to produce a plot with my list of objects. A plot like the
one presented in the link I have provided.
More generally, I would like to use MST as a tool to see if the
distribution is clustered (e.g. M11, obviously there is a cluster) or
homogeneously distributed (e.g. IC 2391, not obvious that there is a
cluster).
Ultimately, I would like to use the normalised correlation length of the
MST (say, "s") and the mean edge length of the MST (say, "m") to make a
distinction between overall radial density (m/s > 0.8, most-likely there
is a cluster, like for M11) and subclustering (m/s < 0.8, most-likely
there are *no* cluster, like for IC 2391).
I hope I better explained what I wanted.
>
> On Wed, Mar 5, 2014 at 9:42 AM, Steve Boudreault
> <address@hidden>wrote:
>> Greetings,
>> I am new to R, and probably this is a very simple problem with very few
things to do. (Sorry if this is the case.)
>> Here I go ...
>> I have a list of about 1000 objects in the file "list.csv". It looks like
>> this :
>> alpa delta
>> 0.923 -12.345
>> 0.912 -11.222
>> 0.812 -11.763
>> 0.902 -11.923
>> 0.942 -12.001
>> ... ...
>> Alpha is the X axis and delta is the Y axis.
>> I would like to produce a very simple minimum spanning tree plot, like the
>> ones I attached with this message.
>> The informations I got so far using igraph in R implied using an adjacent
>> matrix (e.g. "G <- graph.adjacency(as.matrix(D), weighted=TRUE)").
Maybe
>> I
>> miss something important ...
> I don't know where you collect your information from, but I would
suggest
> to use the documentation of igraph, e.g.
> http://igraph.org/r/doc/aaa-igraph-package.html
>> but I fail to understand why I should need a
>> n x n matrix to make an MST for my list of [X,Y] objects.
>> Can someone tell me how I can produce a simple minimum spanning tree plot?
> Minimum spanning trees are defined for graphs, at least in igraph. What
is
> your graph? A set of points is not a graph.
> Maybe you mean that you have a full graph, with your points on the plane as
> vertices, the weights of the edges are Euclidean distances between pairs or
> points, and you want the MST of this graph.
> But I'd prefer you tell us want you want to do, and then we don't need
to
> guess.
> Gabor
> Thank you very much for your help.
>> Best regards,
>> Steve Boudreault
>> --
>> ============================================================
>> Steve Boudreault, Ph.D.
>> CNRS UMR 8111 / GÉPI
>> Bâtiment 11 - Hipparque Email: address@hidden
>> Observatoire de Paris Phone : +33 (0) 145077868
>> 5 Place Jules Janssen Fax: +33 (0) 145077878
>> 92190 Meudon, France Mobile : +33 (0) 604530082
>> ============================================================
>> _______________________________________________
>> igraph-help mailing list
>> address@hidden
>> https://lists.nongnu.org/mailman/listinfo/igraph-help
--
============================================================
Steve Boudreault, Ph.D.
CNRS UMR 8111 / GÉPI
Bâtiment 11 - Hipparque Email: address@hidden
Observatoire de Paris Phone : +33 (0) 145077868
5 Place Jules Janssen Fax: +33 (0) 145077878
92190 Meudon, France Mobile : +33 (0) 604530082
============================================================