|
ChaSeR: Search and Retrieval from the Chandra Data Archive
Introduction
To run ChaSeR:
chaser
NOTE: If you are not on your own machine and are using ssh to remotely
log into another machine to run ChaSeR, you must explicitly
set the DISPLAY environment variable, as in:
setenv DISPLAY yourmachine:0.0
In general, we advise against running Java applications (this is not
peculiar to ChaSeR) remotely.
The CDA Search and Retrieval interface allows Chandra users to browse
the archive and retrieve any data - public or proprietary - provided
that they are authorized to receive the products.
No specific privileges are required for:
- browsing
- retrieving any products from observations that are publicly
released
- retrieving non-proprietary products from proprietary observations
(e.g., aspect or engineering data)
For the retrieval of proprietary
data products one needs to log in to a privileged account - either a
PI account (good for products from observations for which that person
is PI), or a proposal account (good for products from observations
under a specific proposal). Account names and passwords will soon be
distributed to PIs who will be responsible for passing that
information on to collaborators as they see fit. ChaSeR: does
allow users to change passwords. If you would like the receive
information on your accounts now, or
if you need additional information, please E-mail
arcops .
Searching in, and retrieving data from, the CDA is a four-step process,
very much like shopping on the web:
- Browse for products
- Select products and put them in your shopping cart
- Check the contents of your shopping cart
- Place your order
Steps 1, 2, and 3 can be repeated multiple times before step 4 is executed.
The S&R application has three interactive screens:
For a graphical representation of the functional flow (PDF format),
click here.
ChaSeR cannot access data products associated with Engineering
Requests (ER). For access to ER data use
RetrievER.
Note that ERs do not contain any data that are of scientific interest
and users may safely ignore them.
The first screen allows the user to specify constraints in order to
find observations that match what (s)he wants. Most of the fields
are self-explanatory and their use becomes clear when clicked on. For
instance, in coordinates one can specify a rectangular region to be
searched or a cone; clicking on the coordinate input fields will bring
up a pop-up window that gives the user that choice.
Source names may be converted to celestial coordinates through the
SIMBAD name resolver.
Names or coordinates may be read from an ASCII file, instead of being
entered on the form, allowing multiple searches to be executed in one
query. When doing so, the other search constraints remain in effect.
ChaSeR is very finicky about the format of the input file
The input file needs to have on its first line the name(s) of the
field(s) that are provided in the input file, separated by a tab if
more than one. Valid fields are: "Target Name"; or "RA", "DEC"; or
"LONG", "LAT"; or "L2", "B2". No leading or trailing blanks. Make sure
that this specification agrees with what is specified on
ChaSeR's screen. The search values are given, each set on a
separate line, and again separated by a tab; the coordinate format
should agree with what is specified on the screen and there should not
be any leading or trailing blanks. The only whitespace allowed are
blanks within a single value field (e.g., a target name or a
sexagesimal string) and a single tab separating two fields on the same
line. Make sure the Search Radius is set.
For example:
L2 B2
123.4 12.3
56.78 -45.678
or:
RA DEC
08 13 36.00 +12 18 00.00
03 47 07.20 -45 40 40.80
or:
Target Name
M81
2MASX J06094582-21402
AM 0702-601
ARP 148
The normal output of the Observation Search is the Search
Results screen. However, the user may request to have the search
results sent to a file.
Press Submit to execute the search.
The results of the search are presented in tabular form. All
observations with status Archived have been publicly released.
Observations may be selected by clicking on them. Control-clicking
selects multiple entries. A contiguous range is selected by clicking on the
first and Shift-clicking on the last. Details and summary images of
selected observations may be displayed by clicking on the
Detail and Images buttons, respectively; summary images
can only be displayed for public observations (status archived).
The user may select certain data product types for selected
observations and add these to the Retrieve List. When the
Add to
Retrieval List button is pressed the selected products
will be added to the shopping cart for the selected observations.
This action may be repeated as often as necessary - with the same or
with different observations, from the same observation search or from
a new observation search. The default selection is all Primary and
Secondary products.
The standard CDA data distribution
consists of the groups Primary and
Secondary data products. However, the user may customize these
packages by clicking on them and deselecting certain products, and/or
by adding items from the Supporting data products group.
Note that the selections are for generic data product types that may
or may not exist for all observations. For instance, all but certain
calibration observations have a Level 2 event list, but a spectrum is
only available for grating observations while images are only produced
for imaging observations. Similarly, ACIS products are
only available for ACIS observations, HRC products for HRC
observations. Selecting products that do not exist for a certain
observation does not cause an error to occur; one can, for instance,
safely select all primary products for a mix of ACIS and HRC, imaging
and grating observations in one selection action and receive only the
products that exist for each.
See the Guide to Chandra Archive Data Products
for the full list of data products available.
Please be aware that Supporting data products need to be
retrieved from a different archive server and that therefore
retrievals that include supporting products are inherently slower.
Pressing the View retrieval list button brings up a list of the
current contents of the user's shopping cart.
The user may edit the contents of the shopping cart by selecting items
and deleting them.
Browse selected products will bring up a new window with a list of the actual data
products that are included in the shopping cart, with file names and
file sizes. One should note that the sizes are for the uncompressed
products while transfer will take place in compressed form; on
average, the compression is about a factor 2. Users are well-advised
to check the size of their download before executing the retrieval.
Retrieve selected products will start the retrieval process, either directly to a
user-specified directory (Download via application) or to the CDA ftp site
(Stage in anonymous ftp area). In both cases the data products are
collected in a single tar file. The latter method is
significantly faster; in that case a retrieval report will be
deposited in the file chaser.log in the user-specified directory which
includes the name and location of the tar file. If the size of the
file exceeds 2 GB it will be cut into pieces no larger than 2 GB.
The files will remain in this location for 24 hours. Use the log file
for cutting and pasting the tar file name; on most platforms this does
not work well from Java applications directly.
The tar file will contain one directory per ObsId.
If the user is retrieving public data, that is all there is to it.
However, if the user requests proprietary data, (s)he will be asked
whether (s)he wants to log in to another archive account
before the retrieval is executed. The Specify new login and
password button may also be used to change the archive account that is in
use, for instance, for retrieving another set of proprietary data.
Return to search results allows the user to add more items to the shopping
cart or to start a new retrieval process (in the latter case, empty
the shopping cart before returning to results).
See the Guide to Chandra Archive Data Products
for the list of data products available.
To download and install ChaSeR, go to the
download page .
This page offers download packages for various platforms, with and
without Java. Installation simply consists of executing the
downloaded file. If you are not entirely sure of the Java version
that is installed on your machine, we suggest that you take the
download version that includes Java and to install it together with
and specifically for this application. On Unix-like platforms,
the recommended method is to create a symbolic link in one of your bin
directories to "chaser" in the installation directory: do not move the
executable.
In addition, system administrators at sites where the run directory is
not the same as the install directory need to edit the directory paths
lax.dir and lax.root.install.dir in the chaser.lax file to reflect the
run directory path. For instance, if the installation is
done on the file system /foo in /foo/opt/CDA_SR, but /foo/opt is
mounted on user systems as /opt, then the directories will show as
/foo/opt/CDA_SR in chaser.lax and should be changed to /opt/CDA_SR.
Note: Window managers like twm and tvtwm not
only present a problem for running ChaSeR, but may also affect
the installation. See also the FAQ.
In general,
if you choose to download ChaSeR without our bundled Java installation,
you will need to have Java installed on your machine or accessible
from your machine. The Java Virtual Machine (VM) or JRE (Runtime
Environment) version that you need is Java 1.3.1. This Java version
supports the new Java Swing components. Any version earlier than 1.3 will
not work with ChaSeR. The bin directory for your currently
installed Java 1.3 also needs to be in your PATH environment variable
for InstallAnywhere to automatically find it.
Known bugs and current limitations:
- V&V reports cannot be viewed or retrieved.
- There are no links to the literature at the ADS.
- Retrieval of pipeline logs is not supported.
- It does not work well with older window managers like
twm and tvtwm; this is true for all Java applications
and cannot be fixed since nobody is maintaining these window managers
anymore. See also the FAQ.
- If the Cancel button is pressed too quickly (a few seconds)
after a browse or retrieve is started, the screen will freeze.
- Selecting a range in galactic coordinates works reasonably well for
small rectangles, as long as one stays away from the poles. Large
ranges and ranges covering the polses may give surprising results.
- For certain ObsIds retrieval of secondary and/or supporting products
(e.g., aspect solutions) may not return all products that are required.
For feedback and support, go to the
CXC Helpdesk.
Please include as complete information as possible.
This page is maintained by the Chandra Data Archive Operations
group
|
