Документ взят из кэша поисковой машины. Адрес
оригинального документа
: http://www.stsci.edu/ops/pc_procedures.html
Дата изменения: Wed Mar 23 18:15:24 2016 Дата индексирования: Sun Apr 10 13:03:18 2016 Кодировка: |
This document is intended as an introduction to HST Phase II proposal processing. Many of these procedures and tools/programs are more fully documented elsewhere. In the version of this document on the World Wide Web the names of references are hypertext linked to the existing on-line version (in the postscript version the links appear in italics). Throughout the document, examples use 9999 in place of the actual proposal number.
A knowledge of simple unix commands, and a unix editor (like emacs or vi) is assumed in this document. Various system files need to be created or updated to enable you to run the programs used in these procedures. Please follow the Account Set Up Procedures before attempting to follow these procedures.
There is a list of who to contact on the web with questions about various topics.
Revision B - Changed sections: 2.3 (no $'s in unix password), 4.4 (removed information about FOSSIM), 11.4 (added pointer to Duplication Procedures), 14 (made SIB review mandatory), 15.2 & 24 (waiving Trans diagnostics), 26 (added reference to SPSS documentation), 34 (added links to reports). 6/29/94 KAP
Revision C - Major changes throughout document. 5/23/95 KAP.
Revision D - Changes related to the delivery of Visit Spike (including USCs), changes to links for compatibility with new replication of Web pages, added new section on single guide star requests, changed the section on hand off to data analysis, added a few tools to the section on "Other Useful Tools". 9/19/95 KAP.
Revision E - Changes related to the delivery of the new Spike GUI, new confirmation chart procedures, loss of the LSs, new guide star acquisition scenarios, and lots of little changes. 8/28/96 KAP.
Revision F - Changes related to new Phase A/B review process, changes in guide star search strategy, changes related to the PEP/ASSIST merger, changes related to the new layout of the Toolbox. 6/17/97 KAP.
Revision G - Changes to web links, additional information about duplication checking, change in the computer that does confirmation charts, change in PC fixup page, add where templates are stored, eliminate transfer. 8/15/97 KAP
Revision H - Many minor changes. Renamed USCC, took out Walkthrough step, added Procheck, renamed a few buttons in Toolbox, updated HOPR procedures, added how to add e-mail to history file, added new sections on SNAPs, MTs, CALs, GSs which will be fleshed out later. 1/9/98 KAP
Revision I - Alison reformatted for Framemaker conversion to HTML. Other changes include: added links to new documents on ToOs, MTs, Calibrations, SNAPs, GS troubleshooting, automatic SPSS prep; removed information on manual merging; updated path names for operational products: added information on maintaining this document. 6/18/98 KAP
Revision J - Tony gave me two additions: What to do with H&S errors and how to set a visit flight ready from the command line. 7/17/98
Revision K - updated procedures to include the new guide star system. Updated lost links, merged with the SPSS prep procedures AKS 11/24/98
Revision L - updated to include new CS procedures and other changes.
Revision N - updates to schedulability
Revision O - added isql standards in an Appendix 2/14/01 ASV
Revision P - added APT 4/28/06 ASV
Revision Q - fixed broken links and updated old information 03/31/14 BAP
During the annual Phase I process, proposals from General Observers (GOs) are reviewed by the Telescope Allocation Committee (TAC). Proposals are selected for Phase II implementation, and an amount of time (in orbits) is assigned to each proposal. The Principal Investigators (PIs) for the selected proposals are notified and provided with a peer review of their proposal, the number of orbits allocated, the Phase II Proposal Instructions, and updates on instrument status.
In parallel to the PI notification, the Phase I proposals are posted on the internal Phase II proposal information pages, and the Phase I data is loaded into the database. At the same time, each program will have a program coordinator (PC) and possibly a contact scientist (CS) assigned to it. CS assignments are made for all Target of Opportunity proposals (ToO), moving target proposals, and Director's Discretionary proposals (DD), or at the request of the PI or PC. These instructions begin when these assignments have been made.
Calibration proposals, Director's Discretionary (DD) proposals and proposals submitted by Guaranteed Time Observers (GTOs) are not selected by the TAC - so there is no Phase I proposal in these cases. Most of the processing steps are the same, but it will be noted in the instructions when they differ. There are special instructions on line for following special types of proposals: snapshot proposals and moving target proposals.
When proposal assignment is complete, each PC will receive a copy of the notification letter (which includes the TAC Peer Review comments) for their assigned proposals. In the case of a GTO program, you will only be given the GTO's name and e-mail address.
All proposals are stored in the proposal library (PLIB). PLIB is also the name of the software system used to maintain this library. Proposals are stored as "flat" ASCII text files in the RPS2 and APT format.
Access to most of the required tools can be found in one graphical toolbox. The Toolbox is a family of programs written and maintained by Tony Roman. They combine several tedious functions into one button click and a short question and answer session. To bring up the Toolbox type tools at the unix prompt. When it first comes up you will need to choose a submenu by clicking on one of the buttons below "working directory". You can choose any tool within a submenu by clicking on it.
When using the Toolbox, you can either fill in one or more proposal numbers (separated by commas or spaces) on the "list" line of the interface or you can provide a file name on the second line which contains a list of proposals. The working directory listed on the third line is the directory where the Toolbox will look for input files and direct output files. You can change this for the current session, but each time the Toolbox comes up fresh it will assume that your working directory is the one stored with your preferences. You can change this and other Toolbox preferences by clicking on the Preferences button.
This step is only necessary if you are working on a new proposal that was not anticipated (such as a new calibration or a Director's Discretionary (DD) proposal). Most GO, GTO and calibration proposals will already have had a number reserved in the PLIB, so Createpro is not necessary.
To add a new proposal to the PLIB type the following at the unix prompt:
plib create --help
Alternatively you can use the Create button in the PLIB menu of the Toolbox. This button will also prompt you for the TAC allocation (so you can skip the next step). If the file name you give the Toolbox does not exist, the Toolbox will assume that you are trying to reserve a number for a proposal that does not already have a template. It will pop up an emacs editor with a blank template and you can fill in the Title, Proposal_Category, Cycle, PI_Name, and PI_Institution.
Create then does the following:
You will get a message in the toolbox message area telling you that the command ran successfully and what the new phase II ID is.
It is critical that you keep good records because you may be the only person who knows what has been done with the proposal.
The history file is an important place to document what has happened to a proposal. Because the history file is on-line, it is accessible by anyone who may have a question or need to cover for you. It is also useful if you have a less than perfect memory and you have to figure out why certain decisions were made.
The types of information kept in the history file are:
The history files are in /data/implementation/history/ and have the naming convention 9999.history. To create a new history file use the History File button in the Analysis menu of the Toolbox. If you click on the button and the history file already exists, you will be given an emacs editing session on the file.
Exchange Public Folders are used to store all observing program related emails. Each HST observing program has one folder, and each folder is named with the integer ID number of the corresponding HST observing program. New folders will be created by ITSD in large chunks well in advance of when they are needed. Each folder has its own email address of the format HST12345@stsci.edu. Messages sent to these addresses will automatically be filed in the appropriate folders. These folders are visible only to Outlook and Outlook Web Application and not other mail clients. Programs coordinators may read, drag and drop and delete to/from the public folders. Instrument scientists have read-only access to the public folders. It is very important that all email related to a program is placed in the appropriate folder. For more information, please consult the Observing Program Email Folders Procedures.
There are many steps in preparing a proposal for scheduling. This can become confusing when iterating on a single proposal or working on several different proposals at once. A good way to keep track of where you are with a particular proposal is to use procheck. This is a button found in the PC Toolbox. When a field is green, that step has been completed. If a field is red, that step needs to be completed before a visit can be set to scheduling.
This step can be relatively easy with some proposals. The PI may be an experienced HST observer and want to do something very straightforward. But for many proposals you will be answering questions or seeking answers to questions for the PI. You may even need to arrange for the PI to meet with various institute experts.
Only some proposals will have a CS assigned. The CS is an instrument scientist who specializes in the primary instrument used in the proposal. You can send any instrument specific questions to your CS, (or a pass through help desk, si_csreview, e.g. stis_csreview) if no CS is assigned. The role of the CS is outlined in the document: User Support Procedures for Contact Scientists .
It is very important through out the Phase II process for you and the CS to keep each other in the loop on correspondence with the PI. You may make a recommendation that you do not realize effects the science for instance. Conversely, the CS may make a recommendation that adversely affects scheduling.
Send e-mail to the PI for the proposal. Introduce yourself and explain the role that you will play in the proposal's development and implementation. Tell the PI when the complete Phase II proposal is due and the ID number assigned to the proposal. Encourage the PI to use the proposal number in correspondence; this will be especially helpful if you are working on more than one proposal for the same PI. Inquire if there is a Co Investigator (CoI) who will be writing the proposal.
There is a tool called hello-pi that does the following:
Information on when and how to run this tool is usually distributed near the date the PIs are notified of their accepted proposals.
You will receive many types of questions from the PI during the proposal development process. Some PIs will ask you questions about how to achieve certain scientific objectives. If you do not know the answer, these questions can generally be answered by the CS (or by si_csreview). It is critical that you respond to PI inquiries (by e-mail or phone) promptly. If you can not find the answer within a day you should at least respond that you are working on a answer.
Some questions will be about the particular instrument the PI is using. The primary source of this information is the relevant Instrument Handbook. If the answer is not obvious you can seek guidance from the CS, (or email si_csreview).
Some questions will be about syntax. The primary source for this is the Phase II Proposal Instructions. The G version (for the General Observers) is the actual documentation that observers use. The E version (the Engineering version) contains everything in the G version as well as additional options (marked with a black bar in the margin and with special colors) which are only available for engineering and calibration proposals (unless special permission is granted). If you can't find the answer, ask another PC, your CS (or si_csreview).
Some PIs will know exactly what they want to do but have trouble running the APT software. You can help them by being familiar with the APT . When you don't know the answer, you can ask other PCs and/or one of the APT developers.
You may need to run APT yourself on a proposal to answer a PI's question. To run APT just type:
apt &
The APT graphical user interface will come up on your screen. Alternatively, you can select the APT application from the Dock or from the Application folder on your Mac.
At or before the Phase II deadline, the PI should formally submit the completed Phase II proposal. The PI does this by using the "Submit" button in APT. APT submission does the following:
The submission directory (/data/implementation/apt-submission) contains two files for each submission.
The second submission would be marked -002 and so on. Remember to copy the PI's comments (if they are relevant) into the proposal history file. If the PI has made comments or recommendations about the software be sure to pass these along to the apt group at apt@stsci.edu, or Karla Peterson for consideration. If the PI includes questions, they should be answered promptly - just as is done with questions which are sent directly in e-mail.
The first thing to do with the new proposal is put it in PLIB. (It must be in PLIB in order for other people to start discovering duplications with it, and for it to be included in any database queries.) All you have to do now is check in the new version.
Normally in order to check a proposal into PLIB you must first check it out of PLIB. The check out/check in process is intended to help limit the danger of two people working on the same proposal simultaneously. Only one person may have the proposal checked out and only that person has permission to check it back in.
In this case you just want to check in the new version of the proposal without having to first check out the old version. This can be done using the Checkin button in the PLIB menu of the Toolbox. Since you have not checked out the proposal you will be prompted to confirm that you really want to substitute a brand new version of the proposal and whether you want to check in the version in your working directory, or from the Phase II submission directory. The history file will then be brought up in an emacs session so that you can record that you have checked in a new version.
The following directions apply to any change you might need to make to the proposal once it is in PLIB.
In your working directory on unix type:
plib checkout 9999
This will put a copy of the APT file (9999.apt) in your current directory. (Note that it will overwrite an existing file named 9999.apt in the current directory.) You can then edit the APT file using APT.
Alternatively, you can use the Checkout button in the PLIB menu of the Toolbox. It will checkout the proposal into your working directory and pop you into an APT session on the resulting APT file.
Regardless of which method you choose, it is necessary to update the history file (the proposal changes section) to reflect what changes were made and why. Label each set of changes in the history file with your name and the date. The Toolbox will pop up the history file in emacs after you checkout (or checkin in the case of no checkout) a proposal.
If you checkout a proposal and decide that you no longer want to make any changes you can release the lock on the proposal by typing:
plib unlock 9999
Alternatively you can use the Unlock button in the PLIB menu of the Toolbox.
Once you have made the desired changes, in your working directory on unix type:
plib checkin 9999
Alternatively, you can use the Checkin button in the PLIB menu of the Toolbox. For help with plib commands type:
plib help
Once the proposal has been checked into PLIB, APT will run, and pop up a box showing a list of changes found in the proposal and any diagnostics once it has finished.
A high level graphical summary of the PLIB checkin input and output is available in the Diagram of Operational Flow . Note that you must wait for the PLIB checkin to finish before going on to Transformation (Trans). This is because the products that Trans uses as input are created by PLIB checkin.
There are several other nice PLIB commands. Here are a few examples.
The history command shows you each time a particular proposal has been checked out, checked in or even just read. This can be helpful if you have lost track of when a change was made.
plib history 9999
To look at a particular revision of the proposal you can then use the read command to read out a particular revision of the proposal. (Note that it will be stored in the file 9999.apt in your current directory.)
plib read 9999 3
To review the differences between two versions, you can use the tkdiff command. This is quicker than reading out the two versions and using tkdiff, but the results are the same.
plib diff 9999 3 4
If you would like the proposal to be reprocessed but have no changes to make to the proposal, the command is:
plib reprocess 9999
Note that all of these commands are also available from the PLIB menu of the toolbox.
APT is the subsystem that checks a phase II file for problems in syntax. This was actually done automatically when you checked the proposal into PLIB. The resulting diagnostics are summarized in a pop up box and also saved to the directory /data/implementation/apt-report/diagnostics/. If you don't remember the pop-up box, you can use the diagnostics button in the tool box to recall the diagnostics.
Work on a proposal through guide star request and verification should be done as soon as possible. These steps could result in major changes to the program that would affect assignment of plan windows. The rest of the work can be done when it is closer to the time of the proposal's first plan window.
The OPS Toolset (OPS) can be used to run Confirmation Charts (for moving target proposals only), Transformation (Trans), CASM (Constraint Analysis Spike Module), PMDB Load, GS Request, and Spike, Visit Update, and Schedulability.
You can run the OPS toolset from the Toolbox using the OPS Toolset button.
At this point you should run Trans, and CASM which will do the following:
A graphical summary of the Trans and CASM input and output is available in the Diagram of Operational Flow . Explanations and directions for doing these steps individually follow.
The Transformation Scripting Guide is the requirements for the Trans program. In theory, one could create the primary Trans output files (9999*.af) from the primary Trans input file (9999.pp-tdf) using this document. In practice the Trans program does this. The Scripting Guide is the source for figuring out why Trans did something - all the rules are in there (somewhere). The most useful part for a PC is the explanations of the diagnostic messages (errors and warnings). (Also see the Transformation User's Manual .)
Use the Trans button in the Processing menu of the Toolbox.
If you want to run Trans from the command line, here are some examples. Remember to change the proposal number in both places and specifiy how many visits and/or targets you wish to Trans. (The dashes are produced by the word processing program, so leave those out if they appear in the examples below, everything is on one line.)
run-trans ops 9999 ':input "/data/implementation/proposals/" :output "/data/implementation/trans/9999/" :visits "01, 02"' :targets "1, 2"'
run-trans ops 9647 ':input "/data/implementation/proposals/" :output "/data/implementation/trans/9647/" :visits "01, 02"' :targets nil
Output will scroll to the terminal window, and the lisp environment in which Trans runs should exit and leave you back in your UNIX shell when Trans finishes running. You will need to check the trouble report in the Trans directory to see if there were any errors or warnings, see the next section. Running Trans from the command line will not update the database with the dates of the last Trans run. Since the Trans run dates are not updated in the database, you will need to use the "Directory" option under "Determine which entities to load" option. Or from the command line;
pmdb -load -driver=directory
Everything else should proceed normally, until Flight Ready which will give you warnings about the visit change dates and the AF creation dates. You will need to waive these to set the visits flight ready.
The Trans output files are in the directory: /data/implementation/trans/9999/. There are many output files. To get a quick idea of how the proposal Transed, look at the trouble report and the catalog report.
The file 9999-trouble.rpt contains a summary of Trans diagnostics. Figure out which messages require a change in the proposal and which can be waived. There are three types of Trans diagnostics:
Health and Safety Errors: Health and safety errors must be eliminated from the program. If for some reason, there is a desire to override a health and safety error, commanding personnel must be consulted first. If health and safety errors are overridden in Trans, Trans will send mail to commanding personnel notifying them of the override; so it is best to talk to them before Trans does.
Errors: Errors must be reviewed and addressed during processing. In general, no program should be delivered to SPST for scheduling with outstanding Trans error messages. In the event that we must make such a program flight ready (i.e. special engineering programs, new capabilities) we must clearly document in the history file why the error exists and what manual fixups are anticipated to correct it. Unless it is an error that was anticipated (new capability) or has been handled in the recent past and you fully understand the source and remedy, you must receive advice from an SPST expert (like Merle Reinhart) or someone from commanding.
Warnings: Warnings must be reviewed and addressed during processing. No warnings are to be ignored such that the relevant lines are not investigated. Warnings may be waived by the PC but a record must be put in the history file. If you have any questions about the source or action required - ask for help from people in SPST(SMSB), or commanding. Do not simply assume it's probably not a real problem. If warnings are generated that the alignment is too long to fit the visibility, ensure that there is adequate visibility time sometime in the cycle and that you are able to create scheduling windows in SPSS.
Informational Messages: These are for informational purposes only and do not need to be evaluated unless you're looking for an answer that might be provided here. Informational messages generally include identification of overridden tunable constants, actions taken by Trans to maximize efficiency, parameters that have been reset by Trans as they were below/above their minimum/maximum value.
There is a nice on-line version of the diagnostic explanations . This is sometimes handier than the Trans Scripting Guide. It also includes the PI explanations (which are the ones that come up in APT).
The file 9999-catalog.rpt shows how Trans has structured the proposal into the building blocks used in scheduling the telescope. These building blocks are:
It is rare, but sometimes it may be necessary to edit the proposal adding special implementation special requirements and re-Trans in order to obtain the proper structure.
Implementation special requirements are used to request non standard Trans processing of a proposal. For instance if a different alignment or obset structure is needed from what Trans would do by default, this is how you would do it.
There is a list of the implementation special requirements in the E version of the proposal instructions. Simply add the needed requirements to the proposal, and when you re-Trans the stucture should be what you have requested. Inform the PI that you have added these requirements to the proposal and they should not be removed when editing the proposal later on. Also remember to record the fact that you used implementation special requirements as well as the reason for doing so in the proposal history file in the "Progress, Notes, etc." section.
Sometimes it is necessary to override a Trans tunable-constant. These files are named 9999-overrides.lisp and are kept in the /data/implementation/proposals/ directory. See other files in the proposals directory for examples. All files should contain only lines of the form:
(Override <constant-name> <new-value>)
Some commonly used commands are used for moving target proposals and include:
(Override MOSS-OVERRIDE-CYCLE-START "1 Nov 2001")
(Override MOSS-OVERRIDE-CYCLE-END "1 Aug 2002")
(Override Max-Moving-Target-Scan-Time 3200)
The Constraint Analysis Spike Module (CASM) is used to create the links assignment file (AF). This file contains all the timing and orient links between visits. CASM will make the links AF by default unless there are no links. In order to just make the links AF from the command line type:
generate-linkaf -proposal 9999 -spf
CASM can also be run using the links button in the toolbox. (The spf is the CASM scheduling parameters file, which is like the control file for Spike.)
The submission system sends a generic acknowledgment to the PI that his/her proposal has been received. Many PIs prefer to receive a brief email from the PC saying that the program has been received. If you decide to send such an email, it should be done within a day of receiving the submssion. Some PCs may decide to wait until an initial processing of the program before sending an email to the PI. This allows the PC to send a more informative message regarding any problems with the proposal. This is a personal preference.
Implicit proposal verification probably occurred if you were heavily involved in the proposal development, but this is the more formal process that should take place after the PI has put the finishing touches on the Phase II submission. While verification can be run after the program has been transed, most PCs prefer to wait until the program has been processed through visit update.
The process given below has been streamlined by the addition of the Verification button to the Processing menu of the Toolbox. This button provides most of the information described below in a single display. You are also given the opportunity to record the result of your checks in the same panel.
The proposal listing is a more readable version of the information in the APT file. You can see this version on the web by clicking on formatted listing.
Now read the proposal carefully. Compare the intent of the PI as described in the abstract, observing description and other text with the actual target, visit and exposure information given in the proposal to make sure that they are consistent.
Reread the Phase I version of the proposal and TAC Peer Review comments. Confirm that the proposal has not changed significantly from what the TAC approved (targets, instruments, spectral elements, etc.) and complies with TAC instructions. If a recommendation was not followed you should seek guidance from the CS for the proposal (or si_verification). The verification display from the Toolbox will provide the observing summary and the TAC comments for the proposal, but not the entire Phase I proposal.
The TAC allocated a specific number of orbits to each proposal. The estimated number of orbits a proposal will require is calculated by Trans and populated into Assist. Obviously the number of orbits used must be equal to or below the TAC allocation. The verification display from the Toolbox will provide a comparison of the TAC allocation and the estimated number of orbits from Trans.
The duplication check confirms that the proposal does not repeat the work of other proposals (and therefore infringe on someone's proprietary data and/or waste spacecraft orbits) or future GTO proposals (GTOs have strict rights to observations they have reserved). The GOs are required to do their own duplication checking (using the StarView interface to the Archive) before requesting time in Phase 1. Future GTO observations are reserved in a catalog that is also available through StarView.
All of the following duplication criteria must be met:
The process of duplication checking is possible because there is software to search the Assist Database for you. (The Assist database contains the target and exposure information of all the proposals in PLIB.) The check is of proposals that still have proprietary data, not all proposals.
duplication --help
The results are stored to your current directory in a file called: 9999-duplication.rpt. Alternatively you can use the Verification button in the Processing menu of the Toolbox. The GTO reservation catalog is converted to proposal format and stored in Assist also (the proposal number in the 500's). Obviously you will find duplications between a GTO proposal and the reservation catalog for that GTO's team; there is no need to clear those with the CS (or si_verification).
Once the proposal has been examined and found to conform to its Phase I and TAC comments, be within its TAC allocation, and not be in duplication with another proposal, the Verification is complete. You record the fact that verification is complete in the Assist database by simply answering yes to the question of whether the proposal passes verification. This flag is used to signal to the CS/IS that the proposal is ready for the CS review.
Load the proposal into the PMDB (Proposal Management Data Base) by typing:
pmdb -load -type=<prop|proposal_info|su|link|target> <entity ID>
pmdb -load -type=prop 09999
pmdb -load -driver=directory -type=SU 0999901
Alternatively this can be done by using the PMDB Load button in the Toolbox, or through OPS.
PMDB Load loads Trans assignment files, CASM assignment files, (MOSS Chebychev files, and MOSS window files if necessary) into the PMDB, if these files have not been loaded before, or if the files have been updated since the last load. It also sets the scheduling unit general data (SGD) and guide star request windows.
If your proposal requires ISQL (most do not) now is a good time to write it. ISQL files get run at different times depending on what database fields are being updated. See Appendix D for details.
Obset update calculates HST pointing data. Obset update is usually run as part of the GS Request button in the Toolbox (see next step) but can be run seperately by using the Obset Update Only button in the Toolbox's GS Request pop-up window, or from the command line by typing:
pmdb -update -type=obset 09999:01
Requesting guide stars can be accomplished by typing:
gs_request <obset ID>
gs_request 09999:01
or by using the GS Request button in the Toolbox. If you use the command line, the obset update must be run manually, however, if you use the Toolbox, obset update is run for you. On line documentation is available.
Occasionally a visit will have insufficient guide star pairs returned. This will be obvious when reviewing the suitabilities in the next major section. Lack of guide star pairs can be due to sparse or crowded star fields. Scheduling limitations such as BETWEEN or ORIENT special requirements only make the problem worse because these limit the spacecraft rolls that can be used.
One option is to try looking for single guide stars. Changing to single guide star tracking degrades the pointing because roll is maintained by the gyros instead of a second guide star. In many cases it is acceptable to use single guide stars instead of pairs. Review with the CS (or si_csreview) and PI the roll errors associated with single guide star pointing to see if they are acceptable.
To request single guide stars, use the implementation special requirement GS ACQ SCENARIO SINGLE on the first exposure in the visit (or obset if there are multiple obsets). Now checkin, rerun Trans, reload, and re-request guide stars.
For more information on solving difficult guide star problems, see Procedures for Resolving Guide Star Problems. on line.
Bright Object Alerts (BOAs) are generated during GS request, and can be viewed by typing:
gs_request -status 09999:01:01
For FGS proposals, Denise Taylor and/or Bill Januszewski evaluates the BOAs. If you have tried to set something flight ready, and the waiver has not been filed, contact Denise or Bill.
For COS, STIS/NUV-MAMA, STIS/FUV-MAMA and ACS/SBC proposals all BOAs are cleared at the time of the CS review. The review sets the waiver in the database. If the proposal is re-loaded after the CS review, but has not changed in a way that would effect the BOA the PC can reset the Instrument Review complete date in the database Instrument Reviews tool.
Spike is the tool that we use to determine the schedulability of visits. It can be run from OPS or through its own graphical interface. Spike creates the CASM Visit Description (CV-Desc) files which store the suitability information that can then be reviewed using the Spike GUI (see next section). The CV-Desc files for all the proposals are also used to create the LRP.
Spike can be used to review the schedulability of visits. A high level treatment of the various constraints that effect schedulability is available on the Web.
Use the Spike button in the Analysis menu of the Toolbox to bring up the Spike GUI.
Note that this procedure will create the CV-Desc files if they do not already exist or need to be recreated. This process will take at least 10 minutes. If the CV-Desc files are up to date it will only take a few minutes to load the visits for viewing.
To load a proposal into the Spike GUI, click on Proposal, then Load Proposal, then type in the proposal number and hit return or click on OK. Review error/warning messages generated and scheduling suitabilities shown and resolve problems.
Once you have the scheduling units displayed, click on the number of a SU you want to analyze. If you would like to review the suitability components (solar avoidance, guide stars, etc.) in graphical form, click on Expand. If you would like to review the same information as tables of suitable date intervals, click on Suitable Intervals. Either way you will get a new window for that one visit. Other visits can be analyzed in their own windows at the same time.
If a visit cannot be scheduled at any time according to Spike you need to figure out what is causing the problem. The expanded plots are very helpful for this. Changes may be required to the proposal, the way it was Transed, or the guide star scenario.
Two reports that can be very useful are the Target Orbital Viewing Time and Target Orientation reports under the Target Reports menu. Target Orbital Viewing Time will show the amount of orbital viewing available every day and Target Orientation shows when particular orientations are available.
You can print out the post script plots of the suitabilities to put in the folder. Simply click on Reports and then Postscript Plots. The dialog box that comes up gives you the option of printing "Verbose" reports. Without Verbose turned on the plots will show the overall suitability of all the visits. With Verbose turned on the plots will include all the various expanded components. When you click on OK it will print the plots to your default printer.
Now that all the errors that can be caught by software have been dealt with, take a good hard look at the proposal. You may want to keep a check sheet of things that you like to check that software doesn't catch.
A user specified constraint (USC) provides a way to constrain where a visit can schedule without actually modifying the proposal. USCs are stored in the Assist database and can be created and deleted through the Spike GUI.
USCs can be defined in two different modes: combine and override. Note that the mode is associated with the individual USC and not the visit. In combine mode, a USC interval is combined with the other absolute constraints for the visit. In override mode the USC replaces the existing absolute constraints.
Override and combine USCs should be created with caution. Make sure the USC is really doing what you want. Most timing restrictions are better specified in the proposal. One exception is preferences for a time that does not constitute a requirement.
To create a USC, load the proposal into the Spike GUI. Select the Create/Edit USC menu found in the constraint section of the Spike GUI. When the USC menu comes up, choose a visit to work on from the visit menu. Type in a name for the USC in the field called "USC Identifier." Supply a comment about the reason for the USC including your initials. Click on the radio button to select "Combine" or "Override." Then enter the start and end time(s) of the USC in the Cut and Paste Buffer. These should be in the format of start date to end date. Here is an example:
1997.100:00:00:00 1997.120:00:00:00
1997.140:00:00:00 1997.160:00:00:00
Click on the "Create for this Visit" button. The USC will then be saved in the database and will cause the plan windows to be adjusted accordingly the next time the LRP is updated.
There is now only one formal CS review. This review is intended to find and resolve issues that affect scheduling (and hence the assignment of plan windows). Examples of things that affect scheduling:
Optimization of science (such as spectral element choice, or dithering strategies) is not done.
The instrument scientist (IS) responsible for the CS review submits the review by using the SPAR tool. The appropriate PC receives an email of the review. Make sure you read the review for changes the IS is suggesting or recommending. If there are changes mentioned you should talk to the CS (or si_csreview) and make sure you understand if the change must be made or whether it is something the PI should consider. Suggestions should usually be explained by the CS/IS directly to the PI, especially if it has scientific impact.
The tool which distributes the review also records the date completed in Assist as long as the CS/IS has marked that it is not necessary to return for further review. If further review is necessary then the date of the problem is recorded.
The CV-Desc files from Spike for all the proposals are fed into Spike by the LRP Group which determines approximately when each visit should execute. Spike tries to balance the best time to execute each visit while providing the telescope with a steady and balanced diet of visits.
All science visits (except for SNAPs, pure parallels, and ON HOLD visits) are marked "ready for the LRP" in Assist as soon as both the verification and CS review are marked as complete in Assist. If at any time you decide that a proposal should no longer be folded into the LRP (for example if the visit will be completely rewritten with new targets) you can mark the proposal "not ready" by using the Visit Status option of the Database menu of the Toolbox. (Note that verification of SNAP proposals will set the LRP state to "ready" as well, so if you have already changed it to "not LRP candidate," it will get set back to "ready". You will need to set them back to "not LRP candidate.")
To review the plan windows assigned to your proposals, use the Visit Plan Window menu in the Reports menu in the Database menu of the Toolbox. If no plan windows have been assigned or you have questions about the windows, contact the LRP group by sending e-mail to lrpg@stsci.edu.
Once it has been determined approximately when your proposals will run, you can prioritize your work. Obviously you want to work on the proposals with the earliest windows first. Be careful not to leave difficult proposals until the last minute.
By this point you should know exactly why any outstanding Trans errors or warnings can be waived. If you haven't already, add an entry to the history file in the section called "Progress, Notes, etc." which contains the text of the errors and warnings from the Trans Trouble Report, and explain why the errors and warnings can be waived.
This will calculate scheduling windows, science instrument usage, and constraint and restriction overrides. Must be re-run if any of the above are changed.
You can manually set the SGD windows using the toolbox button. This is the time span through which a visit is prepped for execution. The default value is about a year, but a visit which contains MANY alignments (usually NICMOS) can take forever to process with large SGD windows. You can usually shrink the SGD windows for these observations to be around the plan window span. This cuts prep time, but means that if a visit is not scheduled within its plan window, it will have to be partially re-prepped for a different SGD window (rerun visit update and schedulability). You can use the Visit Update button in the tool box or type:
pmdb -update -type=su <SU ID>
pmdb -update -type=su 0999901
If a visit has Transformation errors, visit update will fail and indicate that the visit has been marked "dead" in the PMDB. If it is appropriate to waive the Transformation errors for the visit, use the visit status button in the PC Tool Box to mark the visit(s) undead, and then repeat the visit update.
It is the PC's responsibility to ensure that the SU(s) will schedule on flight calendars which intersects its plan window(s). The SCHEDULABILITY tool assists you in doing this verification in some cases. It checks that the visits can schedule in their plan windows for visits with moving targets or small plan windows. For fixed target visits with large plan windows, and internals Schedulability assumes that SPIKE's estimates are adequate and sets the Prep Done flag.
Run Schedulability by using the Schedulability button in the Toolbox or by typing:
schedulability <program ID> [visit1 visit2 ...]
schedulability 09999 01
If the visit list is omitted, all visits in the program will be processed.
The following are optional parameters that can be changed:
parameter: -ccl_name=
default: <program ID>sch
function: Name to be used by cclist -save.
parameter: -start_time=
default: current date
function: Start time to be used by cclist -create.
parameter: -end_time=
default: current date + 1 year - a pad to keep within the orbit file
function: End time to be used by cclist -create.
parameter: -verify_windows=<spike_cw | sgd>
default: spike_cw
function: Value passed to calopt -verify_windows qualifier.
parameter: -interval=
default: determined by calopt from constraint_windows_ver_criteria
function: Value passed to calopt -interval qualifier
parameter: -orbit_file=
default: Latest orbit file.
function: Value passed to cclist -create -orbit_file qualifier.
parameter: -handoff
default: disabled
function: calopt -handoff qualifier
parameter: -verbose
default: disabled
function: Display schedulability verbose messages.
parameter: -very_verbose
default: disabled
function: Will cause cclist -create and calopt to be run with
-verbose qualifiers, and will display all LRP replan
messages if a replan takes place.
parameter: -force_calopt
default: disabled
function: Forces calopt to be run. Supersedes all other criteria for
determining if calopt should be run including
su_track.vcw_required field.
parameter: -noreplan
default: disabled
function: Surpresses normal Spike run after calopt in order to replan visits
with new calopt output.
Upon successful completion of SCHEDULABILITY the visits that pass will benoted in the message section of the toolbox. If the plan windows for visits change as a result of schedulability, a box will pop up to inform you of the change.
In some cases, it may be necessary to mark a visit as having completed SPSS prep when it did not successfully pass Schedulability's checks. The Prep Done button in the PC Tool Box can be used to accomplish this.
For moving target proposals, confirmation charts should be run after schedulability. In many cases, the default values can be used. Occasionally, it is necessary to enter specific start time, FOV, and/or exposure. In order to generate a confirmation chart for the target, use the Confirm Chart button in the Processing menu of the Toolbox. The confirmation charts are stored in proposal subdirectories in:
/data/implementation/confirm-charts/9999/
If you need to run confirmation charts from the command line, here is the syntax. If a target number is 'All' or not supplied, all fixed and solar system targets in the program will be processed.
confirm-chart [-chart_time=<SPSS format time>] <program ID> [target number list]
confirm-chart -id 9999 -targets 1 2 5 10 23
The directory in which the confirmation charts are stored is linked to the Proposal information page on the WWW, so in order to have the PI review the charts, simply send e-mail explaining how to retrieve the charts from http://www.stsci.edu/public/propinfo.html. It is best to get the PI to review the charts early (gross errors in pointing affect planning windows). You can send the charts as PDF attachements to the PI if that is easier.
Send an e-mail galley proof to the PI for final review. The galley proof email is a request to the PI to do a final review of the program. The galley proof will include any changes made to the program as well as the plan windows for the visit(s). This is done by using the galley proof button in the toolbox.
Include in the galley proof a current version of the proposal listing and the portion of the history file which includes the list of all changes made to the proposal.
The Processing menu of toolbox has a Galley Proof option. The first time you use this option it allows you to customize a galley proof template for yourself, and then uses it to create a galley proof for the proposal you selected. This tool will also mark the galley proof complete in Assist. (If you want to change your template later, you can edit it directly in the directory /data/operational/preferences/.)
If a visit has any special scheduling requirements that must be manually implemented during flight calendar building, a description of these requirements should be recorded in Assist via the Visit Scheduling Comments option in the database update tool. Only unusual requirements that cannot be expressed in the phase 2 syntax or otherwise automated should be entered here.
To mark a visit ready in the Assist database type:
state-change.tcl 9999 01 status scheduling
Alternatively you can use the Flight Ready button in the Processing menu of the Toolbox.
Visits which have an ON HOLD special requirement when they are initially checked into PLIB will be set to "not active" in the Assist database and this will prevent you from making them flight ready. When the condition for the hold has been met you can use the Update option of the Database menu of the Toolbox to set the visit to "active". Note that ON HOLDs which are added to a proposal after initial checkin will need to be made "not active" by hand (using Update).
At any point during this process you could receive a change request. It is not essentially different from the changes you have made all along in order to resolve problems that you or other reviewers found. The difference is that the change is initiated by the PI to enhance or correct the proposal. Discuss with your CS (or email si_change) if the change varies from the original Phase I proposal or will require much rework. Depending on the type of change, it may be necessary to do duplication checking again. A full treatment of our change request policy is available on the web.
If the proposal has already been marked flight ready, you will need to make sure that it's not on a flight calendar. You cannot yank a visit from a calendar without consulting with the calendar builders. Use the Update program to set the status back to implementation.
There is a suite of reports available that show status information from the Assist database. There is also status information from the replicated portions of the PMDB. These reports can be accessed through the Toolbox.
In the Toolbox: Choose the Reports option from the Database menu.
You may get emails from a few eager PIs just after the data are taken. The first thing to point out to the PI is that all the data may not have been down linked yet. (This can take up to a few days.) If the data have been archived, the PI can contact the archive branch (archive@stsci.edu) for questions about retrieving the data.
Any failure reports are sent to the PI and the appropriate PC and (CS). These failure reports are generated by hand after some initial investigation, so they take a few days to show up.
If an observation is found to be degraded, the PI may want to file an HST Observation Problem Report (HOPR). This can be done by the PI directly using the HOPR submission page on the Web. The request for a repeat of all or part of the observation will be reviewed by the Telescope Time Review Board (TTRB). If it is approved, the PC will need to create a new visit, make any needed changes and prep it.
It is very important when dealing with a PI concerning a HOPR that you not accept responsibility for the failure - even if you feel that way. The TTRB will make the decision and you should not make promises to the PI that you think the request will be approved.
If the HOPR is approved and the PI gets to reobserve the target, you will need to create a new visit in the proposal. Here are the steps.
Why did the visit fail? If there was a problem with the visit that must be corrected make sure you understand what needs to be fixed and who will do it.
The usual naming convention for a HOPR repeat is to add 50 to the visit that must be repeated. So visit 6 would be repeated as visit 56. If you use this convention it will be more clear to you and others where this new visit (that puts the proposal over its allocation) came from.
Use the Link Failed To Repeat Visit button in the Database menu of the Toolbox to link the new visit that you made to the visit it repeats. This will allow the Proposal Information web page to show the link between the two. There is a tool which runs weekly which will send you e-mail about any approved HOPRs that do not have such a link.
The little visits in a snapshot proposal are used to fill the holes on a calendar. There are instructions for processing snapshot proposals available on line.
When a PI wants to investigate unexpected or unpredictable transient phenomena, they apply for ToO time in Phase I. There are instructions for processing target of opportunity proposals available on line.
Solar system objects like planets, moons, comets and asteroids also require special processing. See the instructions for moving target proposals available on line.
There are many useful tools that are available, but were not discussed in the main body of this document:
500 series proposals - Each PC will be given at least two proposals in the 500's as their personal testing area. You can check a test proposal in and process it all the way into the PMDB and not touch the operational proposal. If you forget what your test proposal's numbers are you can get a list by typing isql at a terminal prompt and then:
1> select prop_id,pc from prop_track where prop_id > 499 and prop_id < 600
2> go
date convert - This is a graphical data conversion tool which can convert between calendar date, day of year, and Julian date. To bring it up use the DATE CONVERT button in the Other Tools menu of the Toolbox.
Orbital Visibility - This is a Spike tool which is available from the Reports menu of the Spike GUI. It allows you to input an RA and Dec and a time span and gives you a daily report of visibility for that target. (This is the same report as Viewing Time under the Target Reports menu, except the target does not already have to be in a loaded proposal.)
SMS Build Status - Available through the Other menu of the Toolbox, this utility shows the status of all the calendars currently being built by the schedulers and who is the calendar supervisor.
tkdiff - This is a slick tk graphical interface to the unix diff command. Given two files to compare this utility will display them side by side highlighting the different sections one at a time.
xprint - This is a graphical printing tool and is available by typing xprint at the unix prompt, or use the Xprint button in the Other menu of the Toolbox.
wisql - This is a graphical tool for creating a customized query and can be invoked from the command line. PCs generally should be able to get what they need from the Reports menu. If you need to do a "unique" query you can use wisql, but you should use it with care because database edit features are available. If you find that you are doing the same "unique" search many times, it may be time to create a regular report for the purpose.
Test scheduling is automatically performed on all regular, external (non-SNAP) proposals via the Schedulability tool (CALOPT sub-function). For some proposals (moving targets, difficult internals, proposal with special commanding), it may useful to create a test calendar, manually test schedule the SU(s), and review the calendar output files.
Following are the basic steps required.
$ cclist -create [-nolink] <SU ID>
Respond to prompts for calendar start and end times.
$ cclist -save <ccl_name> [-description="description of C&CList"]
Note: CCL's are filed in the directory: /data/implementation/test-calendars/
The test C&C List should be created for the number of weeks in which you want to test a visit. Accurate orbit file data currently extend for a 10-week period, while less accurate extended orbit files may have been produced. If no orbit file covering the target time frame exists, then depending upon a review of the proposal, generate an extended orbit file or postpone processing until an accurate orbit file is available. Linked SUs with links > 7 days may be added to the C&C List by using the additional qualifier -nolink. The SPSS software will ignore the linkage requirements allowing you to test schedule as if these SUs are non-linked.
Test calendars must contain alphanumeric characters and hyphens ONLY and must be 9 or fewer characters in length.
See the SPSS on-line help for full details on the SPSS commands that can be used during the test scheduling process:
$ cadlist -addsus <SU list> [opt. param.]
Add candidates to candidate list. SUs are first added to a calendar as "candidates" (with no scheduling date). Use this command if you didn't add the SU(s) to the calendar during the cclist -create step.
$ candlist -delsus <SU list>
Delete candidates from Candidate list.
$ calendar -addcand <0ppppss> <place> [-verbose]
Add SU to place on calendar. SU must already be on CCL as a candidate.
The syntax for <place> is as follows:
"0xxxxss, path, gap, t0, t1, order"
0xxxxss is an SU that is already scheduled. You are requesting that your current SU be scheduled on the calendar after 0xxxxss.
path is used with interleaver SUs, and you'll never need this. gap is also used with interleaver SUs, and you'll never need this.
t0 is the start time of the observation.
t1 is the end time of the observation.
order is either A or B. Its explanation is beyond the scope of this document. Always use A . If you're curious, look up its explanation in SPSS on-line help calendar -addcand.
You should only need 0xxxxss, t0, t1, A .
The quotes and all commas must be included in the command.
The -verbose option is useful when you are trying to understand why a visit will not schedule. Diagnostic messages will be printed to the screen as the tool is trying to schedule the visit.
$ calendar -addcand 0652203 ",,,1996.125:01,1996.127,A"
Add SU 0652203 on or after 1996.125:01 and before 1996.127.
$ calendar -addcand -verbose 0672504 "0612303,,,1996.200,,A"
Add SU 0672504 after 0612303 and on or after 1996.200. Give all diagnostic messages.
$ calendar -delcand <0ppppss>
Delete SU from Calendar (puts it back in Candidate list).
$ calendar -fbp <0ppppss>
Find Best Place on Calendar for Candidate SU.
$ calendar -fbc <place>
Find Best Candidate in Candidate list for particular place on Calendar.
$ calopt -interval=86400 <ccl_name>
This command will attempt to schedule all SU(s) in the Candidate List on each day (i.e., at an interval of 86,400 seconds) of the Calendar. Results will be stored in file called <ccl_name>.usr. (The CCL must be saved before running this tool.) A smaller time interval may be used if necessary. Review the .usr file to see how frequently and how efficiently the SU(s) scheduled. If any large gaps in schedulability resulted, investigate why and determine if any additional steps can be taken to reduce or eliminate them. Be sure that all Special Requirements are satisfied.
pw (Planning Windows):
Graphical tool which is useful for determining when an SU might schedule, given constraints such as SAA, dark time, etc. Does not consider Guide Star availability.
$ calgraph
A graphical display of how the scheduled SUs fill the orbit(s). Includes occultations, PCS Acq's and Re-Acq's, SK Slews, and SAA impact. Requires test calendar generation.
$ sgs
A Graphical Interface that can be used to create test calendars, schedule SUs, etc.
$ cclist -delete <cclist_id>
During the test scheduling process, it is sometimes necessary to create different C&C List versions. After the final satisfactory version has been created and reviewed, use the above command to delete the old versions from the database and to delete the associated files that take up a significant number of blocks on the disk. It is necessary to maintain only the empty (baselined) C&C List and the final calendar, which is used in generation of the test SMS.
Be sure no errors occurred during the scheduling and that all Special Scheduling Requirements have been satisfied. Review with appropriate parties (e.g., CS, PI, Commanding personnel), if necessary.
This is a list of the PFORMs that the PC may use in modifying a proposal in the PMDB. They are grouped by the major element that they effect (e.g. obsets, targets, guide stars). The most commonly used PFORMS are marked with an arrow. Note that PFORM's can only be used on the VAX machines, not the Alpha's.
PF NAME OF PFORM RELATIONS INVOLVED
-- ------------- ------------------
PROPOSAL
--------
PCM COMMENTS DATA qppcomments
PCP CO-PROPOSERS qpcoproposer
PGD GENERAL DATA qpcoproposer,qpersonnel
SCHEDULING UNIT
---------------
SCM COMMENTS DATA qsucomments,qsbranching
SCR CONSTRAINTS & RESTRICTION OVERRIDES qsoverrides
SDP BRANCHING DEFAULT PATH qsbranching
-> SGD GENERAL DATA (gives SGD windows) qscheduling,qsbranching
SIU SCIENCE INSTRUMENT USAGE qsinstrument
OBSET
-----
BCM COMMENTS DATA qbcomments
BCR CONSTRAINTS & RESTRICTION OVERRIDES qboverrides
-> BGD GENERAL DATA (gives BGD windows) qbs_obset
-> BWD WINDOWS qbwindows
ALIGNMENT
---------
ACM COMMENTS DATA qacomments
-> AGD GENERAL DATA qalignment
-> AID SCIENCE INSTRUMENT DETECTORS qasi_states
AOF REUSABLE OFFSET qaoffset
-> APD POSITION DATA qaposition
ARD READOUT DATA qreadout
EXPOSURE
--------
ECM COMMENTS DATA qecomments
-> EEL LOGSHEET DATA qelogsheet
-> EGD GENERAL DATA qexposure
EPF PODPS FLAGS AND INDICATORS qexposure
-> ESP SCIENCE INSTRUMENT NAMES AND VALUES qesiparm
PARALLELS
---------
PAD PARALLEL ATTACHMENT DATA qparallel
LINK SETS
---------
-> LST TIMING LINK SET DATA qslink_info, qslink_timing
-> LSO ORIENTATION LINK SET DATA qslink_info, qslink_orient
TARGET
------
TCD CHARACTERISTICS DATA qtargets,qalignment
TCM COMMENTS DATA qtcomments,qalignment
-> TEF EPHERMERIS FILES (NON-JPL) qtephemeris
TPS POSITION SPECIFICATION DATA qtargets,qalignment
TRD ROTATIONAL DATA qtargets
TSY SYNONYM NAMES DATA qtsynonyms,qalignment
APERTURES
---------
APR SI APERTURE DATA qaapertures
GUIDE STAR
----------
-> GAQ ACQUISITION CRITERIA qbacqs
GPS PAIR SELECTION DATA wggs_pairs
GSA ACQUISITION SCHEDULING DATA wgacquis
GSP PARAMETERS wggs_pairs,wguide_stars
GSS PAIR SELECTION/ACQ ORDER qbgsp_select,wggs_pairs
GSW OBSERVATION SET WINDOWS wgreswnd
IF NAME OF IFORM RELATIONS INVOLVED
-- ------------- ------------------
IAC INSTRUCTION ACTIVITY DESCRIPTION qexposure
*** Some keys have 2 functions:
HIT KEY TO USE TOP COMMAND
USE GOLD, KEY TO USE BOTTOM COMMAND
_______________________________________
| | HELP | RETRIEVE | CLR FLD |
| | | | |
| GOLD | | ABSOLUTE | | Hitting GOLD, HELP will
| | KEYPAD | APPEND | CLR FORM| bring up this diagram on
|______|________|___________|_________| your screen.
| |SECTION | | |
|SCAN | | INS CHAR | REMOVE |
| | SWITCH | | |
|ZOOM | FORM | INSERT | UNREMOVE|
|------|--------|-----------|---------|
| | | | DEL CHAR|
|FOR- |REVERSE | | |
|WARD |PRIOR | GLOBAL | INSERT |
| | APPL | CLR FORM | HERE |
|------|--------|-----------|---------|
| | | COUNT | COMMIT |
| LIST | | | (Append |
|FORMS |DEL EOF | SET | or Mod- |
| | | DEFAULT | ify) |
|---------------|-----------| |
| ADVANCE LINE| QUALIFY | |
| | | DELETE |
| EXIT | RESET | |
|--------------------------------------
Other useful keys:
RETURN Move to start of next field
TAB Move to start of next field
BACKSPACE Move to start of prev field
DELETE Rubout a character to left
LINE FEED Rubout a word to the left
CTRL-R Refresh the screen
GOLD, TAB Field after scrolling region
GOLD, BCKSP Field before scrolling region
ESC-F Write to intermediate file
ESC-K Numeric keypad (toggle)
ESC-P Write to attached printer
IMPORTANT: Only PCs who have had specific ISQL training may implement special ISQL. If you have not been trained for this, go get the training or go get help from a PC who has been trained in ISQL. It's usally a good idea to get someone else who knows isql to look over it for errors. For SMOV proposals the reviewer is the PIT team, for instrument commanding, it should be someone from the commanding group.
For the purposes of this document, the SPSS Prepping Process is defined by the following steps:
1. Load the entities into the PMDB
pmdb -load
pmdb -update -type=obset -ready <ppppp:oo:vv>
gs_request
pmdb -update -type=su -pass <sssssss:vv>
Normally, obset update is done automatically as part of gs_request. However, when special ISQL is needed, it may be necessary to run just obset update, execute the ISQL, and then proceed with gs_request.
All ISQL written will be Sybase ISQL. All files containing ISQL will have the extension .isql. The ISQL files will reside in the Trans output directory for the program ($trans_output_dir/<pppp>). File names will be constructed as follows and must be all lower case.
pppp_<step>.isql
pppp_vv_<step>.isql
<step> can have any of the following values:
The program coordinator may choose to use either the single ISQL file for the program or individual ISQL files for each visit which needs ISQL.
The single program file allows ISQL to be run just once rather than once for each visit, but it may have to be edited if a subset of visits need to be reloaded later. If this happens than the ISQL for those visits should be removed from the program level file and placed into individual visit files.
If individual visit files are used from the start, the ISQL will have to be run separately for each visit, but no editing will be needed if a subset of visits need to be reloaded later.
Each ISQL file should start with a comment header delimited by the comment delimiters /* and */ at the start and end. This header should explain what the ISQL is for, when it should be run (i.e. after what step in the prepping process), the date it was written/modified and by whom.
The ISQL itself should be arranged in visit and exposure order within the file with appropriate comments interspersed. The ISQL should not contain a use statement to point at a specific database.
To execute the ISQL, use the SPSS report command. For example, to run the file 7003_load.isql:
report -isql -db=$SPSS_DB -output=/dev/tty $trans_output_dir/7003/7003_load.isql
Here is a good example that demonstrates the above principles (the file name is 7003_load.isql).
/*
This ISQL implements the ITS exposure and fixes a Trans problem
Run after pmdb -load
Written by M. Reinhart, 8-Oct-1996
Updated by G. Chapman, 23-Dec-1996 - Add insert of FGS qesiparm.
*/
/* Visit: 03 */
/* Line: 275 */
/* Setup ITS special instructions */
update qgactinst
set instr_name = "EFGSITS"
where proposal_id = "07003"
and obset_id = "03"
and alignment_id = "04"
and exposure_id = "01"
and version_num = "01"
go
insert into qesiparm
(proposal_id, obset_id, alignment_id, exposure_id, version_num,
si_par_name,
si_par_value)
values ("07003","03","04","01","01",
"FGS",
"2")
go
/* Perform the qreadout fixups */
update qreadout
set eng_data_fmt = "FN",
nom_data_rat = 32,
def_data_rat = 32,
max_data_rat = 32
where proposal_id = "07003"
and obset_id = "03"
and alignment_id = "04"
and exposure_id = ""
and version_num = "01"
go
Note that formatting the ISQL statements in this fashion makes for very easy writing and checking.
Some other examples of isql and the associated Phase II proposal comments are available on line.
When appropriate, entries are followed by hot links to relevant documents and/or web pages.
AF (Assignment File) - This is the main product of Trans. There is one AF for each visit and each target and one for general proposal information. The Links AF is created by CASM in a separate process from Trans. AFs are loaded into the PMDB.
Assist Database - This database is populated from most of the fields contained in the proposals in PLIB, as well as some of the values calculated by the PP for Trans. It also contains most of the tracking information for proposals and their visits. ( Schema )
Alignment - One of the building blocks of a processed proposal, an alignment is one or more exposures taken in the same science aperture at the same telescope pointing.
BOA (Bright Object Alert) - An alert generated by the guide star system when a particular telescope pointing will place a potentially dangerously bright object in a sensitive aperture.
CASM (Constraint Analysis Spike Module) - This is the core of Spike's constraint software and has been configured separately to serve both RPS2 and Spike.
CS (Contact Scientist) - The instrument scientist assigned to support a particular proposal.
Change Checker - Part of the PP which compares the new version of a proposal to the most recent previous version looking for targets and visits which have changed. Eventually the change checker will allow automated reprocessing of only the modified portions of a proposal.
CoI (Co Investigator) - Partners of the PI; additional scientists that are part of the team for a proposal.
Confirmation Chart - A confirmation chart is an image of the sky taken from a guide star survey plate with cross hairs superimposed on the pointing specified. They are used to confirm that the specified pointing is what the PI intended. The PI generates confirmation charts in APT (except for moving target confirmation charts). For an explanation on how to generate confirmation charts in APT, please see the following ( document. )
Create - This tool is used to enter brand new proposals into both PLIB and Assist.
CV-Desc (CASM Visit Description file) - This file is created by CASM to store the suitabilities for a proposal after they have been calculated so that they can be referenced easily to review schedulability or make a long range plan.
DD (Director's Discretionary) Proposal - The STScI Director has a small percentage of the orbits that he can award outside of the TAC process. A proposal that has been approved by the Director using these orbits is call a DD proposal.
FTP (File Transfer Protocol) - This utility is available on both the Unix computers. It allows for easy transfer of files between different computers.
Galley Proof - A proposal listing and summary of changes that is sent to the PI.
GASP (GSSS Astrometric Support Program) - This software consists of functions to retrieve and process data from the guide star survey plates and the GSC. GASP can be used to measure target positions relative to the guide star coordinate system.
GO (General Observer) - A scientist who is awarded telescope time.
GSC (Guide Star Catalog) - All the guide stars used by HST are in the GSC which was constructed from several different surveys. The stars in the GSC are generally brighter than 14th magnitude.
GSSS (Guide Star Selection System ) - The software that chooses appropriate guide stars from the GSC for an observation.
GTO (Guaranteed Time Observer) - A scientist who was part of one of the instrument definition teams and has therefore earned time on the telescope.
GUI (Graphical User Interface) - Programs that you interact with graphically instead of from the command line. Examples of our GUIs (pronounced gooey) are: RPS2, PED, and the PC Toolbox.
HOPR (HST Observation Problem Report) - A HOPR provides a mechanism for observers to inform STScI when a problem has occurred with the implementation, scheduling, execution, or data processing of their observations. It is also the means by which the PI makes a request for failed observations to be repeated. ( Home Page )
IM (Instruction Management) - The commanding group maintains the spacecraft commands that proposals must be translated into.
INS (Instruments Division) - This organization is divided into groups each of which supports the calibration and data reduction for a single instrument. ( Instruments Page )
IS (Instrument Scientist) - The scientists in SSD who specialize in a particular instrument. The person responsible for the CS review since these are (in general) not done by the CS.
LRP (Long Range Plan) - The LRP is an execution plan for a large pool of Visits over a long period of time. Spike uses the CV-Desc files for the available Visits, balances the individual proposal constraints with the overall time and resources available, and comes up with plan windows for each Visit (typically 8 weeks in length).
MOSS (Moving Object Support System) - This is the hardware and software for planning moving target observations. ( Home Page )
OPR (Operations Problem Report) - OPRs are the way to formally file software problems as well as requirements for new capabilities. There is an e-mail mechanism for OPR submission ( How To ), or you can use the web interface .
OPS Tool Set- The Operational Tool Set, runs Trans, MOSS, PMDB Load, GS Request, SPIKE, Visit Update, Schedulability, and Confirm Chart.
OPUS (OSS/PODPS Unified System) - This is the group which performs many of the duties formally performed by the Observation Support System (OSS) and the Post Observation Data Processing System (PODPS). This includes routine monitoring of spacecraft telemetry, receiving data downlinked from the spacecraft and pipeline data reduction for the archive.
Obset (Observation Set) - One of the building blocks of a processed proposal, an obset is one or more alignments that can use the same guide stars.
Phase I - Phase I is the annual process of accepting and reviewing thousands of applications (called Phase I proposals) for time on the telescope. SMO coordinates all aspects of Phase I.
Phase II - If a Phase I proposal is given time by the Director (the TAC actually recommends proposals to the Director) it becomes a Phase II proposal. Technically it is still a "proposal" because it is contingent on successful construction of a feasible observing sequence (the APT file) but we interchange the words proposal and program loosely through the entire post Phase I part of the process.
PC (Program Coordinator) - This person is one of two primary contacts for a PI with a Phase II proposal. (The other is the CS.) The PC will help the PI throughout the observing program preparation process. The PC will also run the software to implement the program after it has been submitted by the PI.
PI (Principle Investigator) - The lead scientist for a proposal is called the PI.
PLIB (Proposal Library) - This is the name for the place where the proposals are stored as well as the subsystem which handles the storage and retrieval of proposals. It initiates the execution of the QM on newly entered or updated proposals.
PMDB (Proposal Management Database) - This is the database portion of SPSS. All elements of a proposal which have been loaded are stored as fields in the relations of the PMDB.( Schema )
Plan Windows - These are scheduling windows (generally up to 8 weeks from beginning to end) for scheduling a visit. They are determined by the LRP group using Spike.
RCAT - A program that retrieves entries from the GSC based on a pointing and a tolerance. (There is now an interface to this program on the web.)
SCS (Science Commanding System) - This is the software that translates the various instrument requests into actual spacecraft commands.
SMO (Science Mission Office) - This is the team that coordinates the Call for Proposals for HST, manages the TAC, informs proposers of their allocations, and reports the results to the Director and NASA.
SMS (Science Mission Specification) - A set of commands that completely specifies all visits that will be executed by the spacecraft during a period of time (generally a week).
SMSB (Science Mission and Scheduling Branch) - This is the team that is composed of the Short Term Scheduling (STS) and Long Range Planning (LRP) groups. ( SMSB Home Page , LRP Home Page )
SOGS (Science Operations Ground System) - This name is used to describe both the software and hardware of the ST ground system. The PMDB is sometimes referred to as SOGS, and the secure computer network that it runs on is called SOGS.
SPSS (Science Planning & Scheduling System) - This is the software that is used to prep a proposal in the PMDB as well as schedule proposals for flight.
SQL (Structured Query Language) - This is the language used to store and retrieve data in a Sybase database. It can be used interactively or imbedded in software. Both Assist and PMDB are Sybase databases.
SU (Scheduling Unit) - An SU (also called a visit) is a group of observations on one or more closely located targets that should be observed together. What should be observed together is determined by the PI using RPS2.
Spike - This is the software that is used to determine Visit schedulability and create and maintain the long range plan. ( Home Page )
TAC (Telescope Allocation Committee) - This is actually a bunch of different committees which meet once a year to peer review Phase I proposals and choose proposals to recommend to the Director for telescope time.
TIC (Trans Into CASM file) - This file gives the CASM module of Spike information about how Trans structured a proposal.
Toolbox - Tony Roman's graphical user interface to most of the PC tools.
Trans (Transformation) - This is the software which determines feasibility and creates the file used to populate the PMDB (the AF). There is a special mode of execution that just determines feasibility of proposals for use in APT.( Home Page , Users Manual)
USC (User Specified Constraint) - USCs provide a way to constrain when a visit can schedule without actually modifying the proposal. USCs are stored in the Assist database.
TTRB (Telescope Time Review Board) - This is a group with representatives from both INS and OED. They make recommendations to the Director's office on policy interpretations for the proposals (is this a duplication, can the PI change targets, will this failed observation be repeated).
Visit - A visit is identical to an SU (a group of observations on one or more closely located targets that should be observed together). What goes into a visit is determined by the PI using the APT software.