Here we provide
some details on how to run models with our public tool.
This application
was designed and built by Gerard Lemson and Laurent Bourges.
You need to
register in order to run models on our server. Please send an email to Gerard
Lemson
(“gerard.lemson”
at domain “mpe.mpg.de”). You only need to ask for “sesam permission”
and specify
your full name and institute.
Follow this link to GAVO, and press
“Go” as shown below:
Choose one of
our default models using the scroll-down list button and press “Choose”.
The
parameters within this page are updated only when “Choose” is pressed.
You can now
modify all the parameters within this page, for more details see section #6.
When you are
ready, press “Submit” as shown below. Your model is now being run on GAVO
servers.
Wait ~30
seconds, “reload” the web-page and click “result.zip” to download galaxy
catalogs.
The “Detail” page
contains results of your run; use the “reload” button in order to update this
page.
Each model
takes ~20 seconds to run, and a few more seconds to write the output data. It
may take
more when the
server is busy, in this case your job will be added to a job queue. You can
check the
queue by
clicking “Job Queue”. Once the run is done and “reload” was pressed, the page
will look
like the
example below. It provides a few plots which compare your model to the
observational data.
Click “result.zip” in order to download full catalogs of galaxies. For more details on this page see section #6.
Files to download The status of your run
-
Each line
includes a set of parameters that are used to fix one quantity (e.g. “f_s”).
-
The quantity name
is the same as is used in the paper.
-
A full
description of each parameter appears at the same line. For example,
the
picture below shows the part which is used to fix the behavior of the SF
efficiency.
The first number on the left corresponds to f_s[0], the second to f_s[1], etc.
-
In some cases
(f_s, f_d, f_c) the user can specify a file name. These files are the ones
used by our default models, and their values are given in the appendix of the paper.
Choose
“-1” is you do not want to use a file; in this case the value of f_s will be
fixed
by
the parameters f_s[0] – f_s[3]. On the other hand, choosing a file name will
force the
model to use numerical values from the file, and ignore the parameters f_s[0] – f_s[3].
-
In the future
we will allow users to upload their own parameter files.
-
Make sure that
the parameter values you use are reasonable. We do not check these values
before running your model.
- All the parameters used by the model are saved to the file: “esam_params”. This file is included
within the archive “results.zip”, but you can download it separately.
-
The results of
the public runs are based on merger-trees from the Milli-Millennium simulation.
This simulation is based on a 62.5
Mpc/h box size, a volume that is 512 times smaller than
the full Millennium simulation used in
the paper. For this reason there might be differences between
results obtained here and the results
shown in the paper.
-
We provide a
few plots to compare the model results to observations. These plots are done in
the
same way as was done in the paper. For more details on these
plots, please refer to the paper. In all
plots the model results are shown in
red.
-
In order to
fully evaluate your model, we recommend downloading the galaxy catalogs. The
*.zip file,
“result.zip”,
includes all galaxies at the following redshifts:
0
0.2 0.4 0.6 0.8 1.0 1.2 1.5 2.0 2.5 3.0 3.5 4.0 5.0
The
most common error is obtained when the efficiencies values are too high.
In this case the error message will look
like the example below. You would need
to check yourself that the values of each efficiency (i.e.
f_s, f_c, f_d, f_e, f_r)
do not exceed the limit of ~100. Such
high efficiencies mean that drastic changes
in the mass component of a galaxy will
occur on time scales of ~10 Myr, which is not
allowed by this public tool.
-
Each user can add
no more than 6 jobs to the queue system.
-
You can view
all your previous runs using the “history” link.
We
will try to keep the full data from each run for at least one day,
after this time only the parameter file, “esam_params”, will be saved.
This
is the first application done by GAVO “Jobrunner” so we need your feedback!
Please
let us know if you have any comments, questions or problems.
Goto: top of this page,
SeSAM main