Version 1.1 (2002-10-21)
The format of the database tables is described by table definition files (TDF). These form the backbone of the QC1 database structure which is otherwise flat. I.e. it has a set of QC1 tables but (at least at the moment) no relations between them. So the TDFs, once finished, completely describe the database structure.
They have a second important purpose: they serve as configuration files
for the qc1 database interfaces. E.g., it is here where the user defines
which keys, in which order, are visible in the qc1_plotter interface.
fors1_bias
fors1_dark
fors1_img_scrflat
fors1_img_skyflat
fors1_img_zerop
fors1_img_fixp
fors1_img_colext
fors1_lss_wave
These are mapped one by one into corresponding TDFs for the QC1 database. These files are then used to construct the database tables, and to provide self-documentation. E.g., in some interfaces, the selected keys are displayed together with the description provided in the corresponding TDF. The TDFs themselves become part of the database.
By agreement with Stefano, the TDFs per instrument can either be provided
as single or multiple excel table (.xls), or as flat ASCII tables.
The set of columns is always the same across all TDFs. It has:
field_name | self-explaining if possible; if keyword_name exists, used a translation |
keyword_name | short version of related FITS keyword or QC keyword, if available; e.g. HIERARCH.ESO.INS.TEMP1.MEAN would translate into ins_temp1_mean |
field_type | varchar/float/integer/datetime |
plot_flag | used for qc1_plotter interface; the qc1_browser also uses this keyword
for default selection. Possible values: empty - not offered in qc1_plotter, unselected in qc1_browser XD - key used as X axis default (one and only one) YD - Y axis default (one and only one) XY - clickable in interface F - clickable for filtering Combinations XDY, XYD, XY, FXY are allowed. |
keyword_type | type of key; used for sorting by qc1_browser; possible values:
general qc1 instrument |
sort_flag | integer number; used for sorting by qc1_browser; within all keys per keyword_type, sorting is done along sort_flag |
length | length for varchar types |
default_value | default values: NULL or specific value |
example | example |
description | used for self-documentation |
The set of entries differs from TDF to TDF. While the set of general
keywords is the same in most cases (if not all), the qc1 keywords are of
course specific. The instrument keywords will again largely be the same
across one instrument mode.
calibfile
pipefile
arcfile
origfile
cdbfile
mjd_obs
tpl_start
civil_date
validity_flag
timestamp
-calibfile: the file name of the master calibration file (our
convention)
-pipefile: the pipefile name of the master calibration file
-arcfile: the arcfile name of the (first) associated raw file
-origfile: the origfile name of the (first) associated raw file
-cdbfile: the cdbfile name of the master calibration file (given on
ingestion, only known for downloaded files!)
-tpl_start: tpl.start keyword from the raw files
-civil_date: format yyy-mm-dd (e.g. 2002-10-22)
-validity_flag: yes by default; no by updating
-timestamp: of last modification
Filenames: For the purpose of being able to cross-correlate (join)
information from the qc1 database with e.g. the raw data archive, it is
extremely useful to have, as part of the qc1 database, information about
the file names associated with the qc1 parameters. Of course this information
is not always present, but we should try hard to retrieve it for historical
data, and to provide for future data by default.
The complete set of TDFs for UVES (echelle mode) is available as
single excel file under the URL http://www.eso.org/~qc/docu/GENERAL/uves_qc.xls.