Sheet1
| Nulling and V2 Reduction Code Status and Plan |
|
CDK 25apr04; update 30jul04 |
|
|
|
|
|
|
|
|
|
|
Key |
Completed |
|
|
|
|
|
Behind schedule (low risk) |
|
|
|
|
|
Behind schedule (high risk) |
|
|
|
|
|
Abandoned / obsoleted |
|
|
|
|
|
|
|
| L1 (V2) |
|
Status |
Date |
Description |
Comment |
|
Archfile read functionality |
Done |
01mar04 |
KI API (by Jen) and wrappers (by CDK, with support from Jen) |
Additional channels may be needed but should be easy to add at this point. Reading of old 1.7 archives needs work if fast-track K-band option is exercised. |
|
Archfile performance verified adequate |
Done |
30apr04 |
Verify adequate read speed using IDL API wrappers for "full" channel list on CARA machine |
Preliminary tests on local machines look promising. Summit testing depends on functional 2.1 API installation. |
|
Quasi-realtime algorithm |
Done |
15jan04 |
Need to verify this with the latest API and wrapper code changes |
This was demonstrated with IDL wrappers some months ago |
|
Calibration procedures for N-band (preliminary) |
Done |
30jun04 |
Have discussed with MMC and ES; no showstoppers found; slightly different from NIR case. |
Lab testing will help, but final method requires sky data |
|
Full-up V2 testing in lab |
Done |
27apr04 |
Run through an observing sequence, including calibrations, in the lab. This will need to be done periodically to verify experimental procedures and code changes. |
First successful test run 27apr04; analysis in progress. More iterations with later code versions and more functional hardware needed. |
|
Analyze 1st fullup V2 test data |
Done |
07may04 |
Derive V2 and other output quantities; verify reading and structure analysis of archfile, in-lab V2 calibration procedure |
|
|
2nd V2 full-up test |
Nuller shipped to summit before this was done |
23jul04 |
Test quasi-realtime reduction of lab data as it comes in |
his test ncluded realtime analysis using v2_l1.pro. It revealed a V2 calibration issue still to be resolved. |
|
N-band L1 coding (overall) |
Done |
15jun04 |
Update K-band V2 code to N-band; add features, documentation |
Can reduce full-up tests in realtime. Some debugging and refinement still needed. |
|
N-band L1 coding (part 1) |
Done |
29apr04 |
Verify archive access against V2 full-up test #1 archfile |
Read all channels, sanity check values |
|
N-band L1 coding (part 2) |
Done |
29apr04 |
Code to construct a list of events (enum channel value changes) with times. File save and load. |
Eventlist is constructed on Pass 1 throughthe archive (for non-realtime operation) if it doesn't already exist; read from file otherwise. |
|
N-band L1 coding (part 3) |
Done |
04may04 |
Identify timetick ranges for various data types |
|
|
N-band L1 coding (part 4) |
90% (workaround installed for API problem reading LDL OPD) |
20may04 |
Averaging and calibration of data; saving of files |
This could actually be counted as 'done' from the point of view of v2_l1 itself, presuming that the problem is in the API as expected. |
|
N-band L1 coding (part 5) |
Done (realtime plots) |
25may04 |
Generation of plots |
Output using code described in 'RT Plots' below. |
|
N-band L1 coding (part 6) |
Done |
10jun04 |
Quasi-realtime functionality |
Implement monitoring of archfile; extend eventlist and recompute calibration groups when changes are found; update plots and sumfile. |
|
L1 test on summit |
Done |
06jul04 |
Run through an archfile of full-up test data on the summit; also try an archfile of fake data growing in realtime to verify quasi-realtime functionality. |
Effectively a check that software dependencies and environment are set up correctly. |
|
L1 code documentation |
Done (1st version) |
30jul04 |
"Cookbook" describing the code is written |
|
|
Update v2_l1 cookbook |
|
20sep04 |
Describe new functionality, calibration scheme; distribute |
CARA folks have expressed interest in this |
|
Move v2_l1 and docs to CVS |
0 |
20sep04 |
|
This probably involves moving my current $KROOT/kss/if/ISC/koresko tree to MSC (msc?) also. |
|
|
|
|
|
|
|
1st Summit Trip |
Done |
19-25jul04 |
Support software and hardware testing on summit; run V2 (and Nulling?) FUTs with latest L1 and RTC code versions; make sure things are ready for Nuller V2 first-light. |
Completed. Was useful. |
|
1st N-band V2 run |
Done |
06aug04 (~01sep04 fallback) |
Goals: Detect starlight on KALI; measure overall throughput; seek and lock on fringes. Various engineering tests. |
Very successful; first fringes detected, science data taken, v2_l1 ran stably for ~6 hrs. |
|
2nd N-band V2 run |
|
24-26sep04 (travel 19-28sep04) |
Goals: As for 1st run but with more complex secondary beamtrain. |
Update v2_l1 calibration scheme and observing sequence based on lessons learned from 1st run |
|
L1 debugging / feature extension |
0 |
continuing |
Make improvements as needed. Ideas: automatic monitoring and switchover to follow new archfile creation; basic L2 processing; overplotting of model predictions. |
Development will continue to accomodate changes in the experiment plan and user feedback. |
|
|
|
|
|
|
|
Option: Accelerated development of K-band prototype |
Done |
Be ready for test 24may04; 30may04 engineering run |
Implement the full system for FATCAT; use it during engineering time to verify functionality. As concieved, it would run in parallel to rtKVis. Depends on full API functionality for v1.7 archives |
Differences between K-band and N-band data are probably modest enough that the verified K-band system can be converted for KALI very straightforwardly (mainly parameter changes) |
|
Decision: Go/No-Go for K-band test during 30May-01June observing run |
Done |
07may04 |
If the API looks functional and coding is going well, shoot for having a realtime K-band V2 reduction package to be tested on the summit during the observing run at the end of May. This code would be written to be easily adapted to N-band data. Otherwise, rely on in-lab and purely virtual testing for development of V2 code for N-band. |
|
|
|
|
|
The accelerated development track was carried through successfully, including a realtime test on the summit (using the stimulus instead of a real star to avoid any possibility of interference with science observing). |
|
|
|
|
|
|
|
| L2 (V2) |
External calibration of V2 data |
Done |
10jul04 |
v2_l1 writes specfiles readable by wbCalib |
Examination of wbCalib documentation indicated that it should be able to process Nuller V2 L1 data; no separate code is needed. |
|
External calibration of V2 data (part 1) |
Done |
20may04 |
Read L1 V2 output ("sumfiles") |
Sync with (depends on) L1 output file writing task |
|
Calibration procedures applied; output files written |
Done |
05jul04 |
External calibration to produce estimated astrophysical fringe visibility |
|
|
End-to-End Testing |
Done |
10jul04 |
Run a dataset through the full pipeline starting with an archfile and ending with final data products |
Will be done a number of times in the course of development |
|
Build/Install info |
0 |
25jul04 |
Instructions for getting the software installed |
This functionality should be provided by nbCalib, which is already installed. |
|
|
|
|
|
|
|
|
|
|
|
|
| L1 (Nulling) |
|
|
|
|
|
|
Archive functionality & performance |
Done |
|
Three steps are essentially the same as for L1 V2 (the channel list is different) |
There's no reason to think it'll break for the Nulling mode, but an API or wrapper bug could surface. |
|
Calibration procedures |
Done |
01jul04 (preliminary) |
This is in progress based on existing (limited) lab data. We actually have procedures derived from theory and simulation but I want to go to the summit with something more realistically demonstrated. |
Preliminary procedures may need to change in response to the behavior of real on-sky data. |
|
Full-up nulling test in lab |
Done (failed due to various hardware and RTC problems) |
|
Retroreflect thermal light through the system to simulate starlight; go through 'observing' steps, generate archfile |
Somewhat more complicated than V2 testing. Currently I'd guess the first one will happen ~5jun04. Repeat as code and techniques develop |
|
2nd full-up nulling test |
Done (mostly worked; some fudging of data needed) |
Done 07jun04 |
Incorporate as much functionality as the hardware can provide at that date; use all channels to be run at the summit. |
|
|
Analyze 1st fullup test data (quick) |
Dropped due to failure of 1st FUT and availability of 2nd FUT |
|
Verify that the data are good and understandable; document them |
|
|
Analyze 2nd fullup test data (L1) |
0.1 (cannot be completed) |
30jun04 |
This goes hand-in-hand with the writing of the L1 Nulling code |
A bug in the realtime software prevented data from being recorded during null measurements. A new FUT is needed to fully test the Nulling L1 code. |
|
3rd Nulling FUT |
Upcoming |
20sep04? |
Another attempt to get a good Nulling FUT archive, this time on the summit before the 2nd V2 observing run. |
I will be in Waimea for this (possibly summit) |
|
L1 nulling coding |
0.75 |
15oct04? (depends on Nulling FUT) |
Code derived from V2 code but with different channels, calibrations, and documentation |
This is going OK. Installing workarounds for missing FUT data. Note that on-sky code will evolve over time to take into account e.g., data from FATCAT. Completion requires a correct FUT dataset, which should be available in time (but after the 2nd V2 run). |
|
Build/Install info |
0 |
15oct04 |
Instructions for getting the software installed |
This will be derived from the "cookbook" written for L1 V2. |
|
L1 debugging / feature extension |
0 |
TBD |
Make improvements as needed |
See coding task above |
|
1st N-band Nulling run |
0 |
~30oct04 |
Goals: Basic functionality of Nulling observing sequence; deep nulls not expected this time out. |
L1 Nulling code to be run in quasi-realtime mode similar to the V2 case |
|
|
|
|
|
|
| L2 (Nulling) |
|
|
|
|
|
|
Read nulling_l1 output files |
0 |
15nov04 |
Files will be ASCII format but somewhat complicated as a reflection of the complexity of nulling calibration |
|
|
Implement L2 calibration scheme |
0 |
20nov04 |
This is the math at the heart of the code |
Note that Nulling L2 calibration consists of estimating reading terms due to internally-calibratable factors (phase and amplitude errors) and measuring external calibration (using calibrator stars) |
|
Write output files |
0 |
30nov04 |
ASCII datafiles (suitable for science analysis) and PNG plots for diagnostic purposes |
The final result is an astrophysical leakage measure with an errorbar |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| RT Plots |
|
|
|
|
|
|
IDL plotting routines |
Done |
|
Simple routines in IDL, to be invoked by quasi-realtime code, to generate truecolor bitmaps containing plots for web interface |
IDL code for generating plots written; debug & test needed |
|
Detailed communication with Claude (need location, image sizes, etc) |
Done |
29apr04 |
|
|
|
Integrate with reduction codes |
Done |
|
|
|
|
Test on summit |
Done |
05jun04 |
Test: plot data from a few channels while archfile is growing |
Many thanks to John Gathright for running the system while I tested the L1 code |
|
Documentation (V2) |
Done |
04jun04 |
Description of the plots and how the code works |
This is just a section of the V2 L1 code documentation |
|
Documentation (Nulling) |
Not started |
15oct04 |
Description of the plots and how the code works |
This is just a section of the Nulling L1 code documentation |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| Run Prep |
For V2 First-Light run |
In progress |
05aug04 |
Construct object list and choose LDL positions; tune v2_l1 for hardware config and experiment design; work expands to fill available time |
Science target list complete; calibration and engineering target list in progress; v2_l1 needs testing against new KALI framerate and calibration / phasor plot issues need work. |
|
For 2nd V2 run |
Not started |
25sep04 |
As for first-light run |
Different hardware config from first-light. v2_l1 is designed to adapt automatically but this has not been tested. FUTs will probably be used to verify it. |
|
Nulling 1st-light run |
0.5 |
25oct04 |
|
Primary calibrators chosen for full nulling observations, but engineering targets are needed for testing, along with easy potential science targets. |
|
|
|
|
|
|
| Data Analysis & Publication |
For V2 First-Light run |
Started |
Before 2nd V2 run |
After the run. Analysis & writeup tasks shared with coauthors (TBD) |
Lots of engineering data. MMC's punchlist has many details. I am supporting MMC, GV, & SC on some of it and doing some myself. Science data were taken and I am working on them too. |
|
For the 2nd V2 run |
Not started |
??? |
Ditto |
Ditto |
|
For the 1st Nulling run |
Not started |
??? |
Ditto |
I'm not as optimistic about the 1st nuller run producing science, but there will be a lot of data analysis needed to refine the instrument and analysis. This will drive changes to nulling_l1. |