Redshift measurement instructions

  1. PANDORA.EZ INSTALLATION

  2. RETRIEVING and UPLOADING SUMMARY FILES

  3. VIPERS SUMMARY FILES

  4. SUMMARY_CHECK_TOOL

  5. SECOND OBJECT IDENTIFICATION


PANDORA.EZ INSTALLATION


Get THIS TAR FILE (username: pngs, password: pngs20133), unpack it and follow the instructions contained in the README and INSTALL_EZ file.

After the installation, download the EZdefaults file and rename it as $HOME/.EZ/EZdefaults.

RETRIEVING and UPLOADING SUMMARY FILES

Login into the VIPERS private pages, using your username and password

Go to the page Spectra->mypointings. You'll find there the list of pointings (and quadrants) assigned to you
On the same page, for each pointing/quadrant you will also see who is the secondary (or primary).

Retrieve the data by clicking on "GET". For each quadrant and pointing, the package consists of
the seq file, the master lamp file and the summary file.
Untar the file in the place you have chosen for vipgi to see them
(typically <whatever directory on your machine> mos/Qx, where x is the quadrant you
will be working on

"measure" the redshifts, keeping in mind the flag conventions as described in the page
spectra->spectra flag in the VIPERS private area.

IF your are the SECONDARY


Once done, send the summary file to the primary measurer , wait for him to come back to you for
reconciliation

IF you are the PRIMARY

get from the secondary his summary file.
Compare your own summary file with the one provided by the secondary. To this purpose you
can use the Summary Check Tool

Solve primary/secondary discrepancies, talking with the secondary (even by phone or e-mail)
till you converge to a solution which is acceptable by both (secondaries: do not be shy!)

If there are second objects which could not be identified by Easylife (id=0000000 and a redshift measured)
or not securely identified targets, provide an identification (see below)

Finally, upload the final seq and summary file through the page Spectra->mypointings

NOTE: during upload, the summary file will be checked  for formal errors (e.g. all objs must have a redshift
and a flag, if flag is 0 redshift must be 9.9999, and many other checks). If the summary file
does not pass ALL checks, the upload is refused. You'll have to fix all errors, and resubmit.
If all checks are passed, it will take sometime to make the upload, (as many other things are
done at the same time). DO NOT CLOSE THE BROWSER TILL YOU GET the successfull messga

You can check that everything is OK by going to the  Survey->status W1  (or Survey->status W4)vipers private page: your pointing hould
 appear in RED

PRIMARIES, don't forget that you are responsible for a fast and carefull measurement.
It is up to you to solicit the secondary to carry out his job.
After a reasonable amount of time (TBD) if your pointing is not finished, somebody will ask you why.

VIPERS SUMMARY FILES

Note that zero orders have been automatically removed from spectra, as part of the standard reduction

The summary files you get will have the following VIPERS specific characteristics
  1. EZ blind has been blindly run on them as part of the standard reduction
  2. The EZ "reliability" flag has been computed as part of the standard reduction
  3. EZ blind results have been compared to photometric redshift, and EZ_flag changed adding a decimal digit according to the following rules (WARNING: these are not the same flags you can find in the final spectroscopic table in the VIPERS database)

FLAG addition
 MEANING
.1
 no photoz available
.2
 spectro z NOT compatible with photoz
.5
 spectro z compatible with photometric redshift, .i.e

minvalue < zspec < maxvalue

where

minvalue=min( photo_z-(1+photo_z)*0.05, photo_z_min )
maxvalue=max( photo_z+(1+photo_z)*0.05, photo_z_max )

0.05 is 2 times the scatter of the comparison between zspec and zphot (0.025).

If you disagree with EZ blind solution, do not worry even if it looks compatible with photoz. Just re-do the measurement and do not worry about photo-z. When summary files are uploaded again, the redshifts will be checked again aginst photoz and the decimal part of the flag updated accordingly; remind that the decimal digit added in this final check follows slightly different rules; see the database tables documentation for details.
Also note that EZ ALWAYS gives a solution, thus EZ_flag 0 objects do have a redshift, which is EZ "best attempt". It could be totally wrong, and it could even be that the redshift is not measurable, or that object is a fake!
Last, but not least, EZ is a program and as such is more stupid than any human being. Use it as a tool, but do not trust it blindly and completetely


SUMMARY CHECK TOOL


Get the script from summaryReconcile_VIPERS and put it wherever you want

Run the script with the command
python summaryReconcile_VIPERS summaryFile1 summaryFile2

where summaryFile1 is the summary file coming from the primary, and summaryFile2 is the summary file coming from the secondary. If the script is in a diffferent directory, prepend its full path name to the script name. NOTE: this script is slightly different from the analogous zCosmos one. Do not mix them
python summaryReconcile_VIPERSsc_W1P040_LRred_M1_Q3_seq.summary_bg sc_W1P040_LRred_M1_Q3_seq.summary_pp

Redshift measurements are considered in agreement if Delta_z < 0.001(1+z). Flags are considered in agreement if the two integer part values are equal or differ by 1.

In the output summary file, the redshift measurement is always the one coming from the first input file to the reconciliation script, as well as the object ID. Therefore, summaryFile1 should always be the one from the primary, who is responsible for II obj identification.
The flag value is decided as follows: if the two flags are in agreement, the maximum between the two values is taken, otherwise the value from the first input file is taken. If the first input file is missing the flag value, the flag is set to -999. Only the integer part of the flag value is kept into consideration (a posteriori, when the reconciled summary is sent back to the data repository, decimal values are computed once again).

SECOND OBJECT IDENTIFICATION

Second objects serenditiously appearing in the slit are usually automatically identified within the automatic reduction.
Occasionally, the automatic identification procedure may fail. In this case, you will find in the comments field some comments.
The table below lists the possible comments, details their meaning and suggests the action to take


 COMMENT
 MEANING
 Action
1
 II OBJ NOT IN CAT Nearest OBJID XX px away
 no obj. found in the photometric catalog near the position of the current obj. The neariest one is at XX pixels along the slit, and its ID is OBJID. It is probable that the current object is a fake
 If no redshift can be measured, the object can be considered as a fake: z=9.9999 flag=-99



If you measured a redshift, this is an unidentifed second object. Use the II object identification tool: if you find an object, change ID and mag in summary file, flag will be 2x. If no counterpart can be found, ID and mag remain 00, flag will be -2x
2
 TARGET? ra xxxxx dec xxxxxx
 the coordintaes of the target obj (as from DB) are too different from the current obj coordinates. It is probable that the current object is NOT the target, could be another object or a FAKE
 If no redshift can be measured, the target is not detected, and the object is a fake:z=9.9999 flag=-10



If you measured a redshift, the target is not detected, and this is a second object. Put z=9.9999 flag=-10. You do not need to identify such object.
3
 TARGET ra xxx dec yyy this obj zz  kk px away
 There are 2 objects very next one to the other, the program could not decide which is the target. You'll have the same comment for another object in the slit
 One should be the target, the other one a second object (or a fake). Use the II object identification tool to decide which is which, then edit the summary file accordingly.
4
 II OBJ
 this is a serenpitous object in the slit, which has been already currently identifed
remember to put flag 2x

SECOND OBJECT IDENTIFICATION TOOL

Spectra->Second Object
username: obsVipers
password: ven0m (NOTE THE ZERO)
You will get a list of objects around the target known to the database. The first line is the target, the following lines are objects in the surroundings. You will also get the CFHTLS_I band image around the target. Check the list and/or the image to  identify the second object. In doing so, remember
  1. Images come with the SAME orientation as VIMOS raw data
  2. Objects in slit are numbered (usually, but not always) in increasing order. So  that (usually, but not always) object 1 will the leftmost in the slit, object  2 more on the right
  3. Slits are oriented along the E-W direction. Second objects should have the same declintaion, within 1 arcsec (unless the serendipitous it is a very bright object the outskirts of which fall in the slit)
  4. In the summary file comments, you have some infos to help you out. For example, in case 3 above, you have the distance ALONG RA (in VIMOS pixels, 1 px=0.205 arcsec) between the target and the current object. Check the list and image to disentangle which is which.