|
1 | 1 | Roadmap |
2 | 2 | ======= |
3 | 3 |
|
4 | | -- Fitting ESIS Distortion |
| 4 | +- ESIS Distortion Model |
5 | 5 |
|
6 | | - - Pull merit functions and optimization routines out of notebooks and into codebase. |
7 | | - - Save results of optimization runs into format that can be read into an instrument instance. |
8 | | - - From distortion fits build ``InterpolatedSystem`` class that fits a desired polynomial |
9 | | - to raytrace for faster mapping from source to detector. |
| 6 | + - Pull merit functions and optimization routines out of notebooks and into codebase. |
| 7 | + - Save results of optimization runs into format that can be read into an instrument instance. |
| 8 | + - From distortion fits build ``InterpolatedSystem`` class that fits a desired polynomial |
| 9 | + to raytrace for faster mapping from source to detector. |
10 | 10 |
|
11 | 11 | - CTIS Package |
12 | 12 |
|
13 | | - - Calculate weights from interpolated system |
14 | | - - Transpose weights for backward operation |
15 | | - - Define standard interface for all inversion methods |
16 | | - (forward, backward, data, merit, inner/outer loop filtering? |
17 | | - to be ready for: |
| 13 | + - Calculate weights from interpolated system. |
| 14 | + - Transpose weights for backward operation. |
| 15 | + - Define standard interface for all inversion methods. |
| 16 | + (forward, backward, data, merit, inner/outer loop filtering? |
| 17 | + to be ready for: |
18 | 18 |
|
19 | | - - CNN |
20 | | - - MART/RL (other iterative methods) |
21 | | - - Possibly sklearn linear regression? |
| 19 | + - CNN |
| 20 | + - MART/RL (other iterative methods) |
| 21 | + - Possibly sklearn linear regression? |
22 | 22 |
|
23 | | - - Figure out the best way to store results in a common object so we can examine convergence/progress and compare methods easily |
| 23 | + - Figure out the best way to store results in a common object so we can examine |
| 24 | + convergence/progress and compare methods easily. |
24 | 25 |
|
25 | 26 | - Data Synthesis |
26 | 27 |
|
27 | | - - Forward model from VDEM to detector. |
28 | | - - Generate synthetic ESIS-II images from simulations. |
29 | | - - Add noise "correctly". |
| 28 | + - Forward model from VDEM to detector. |
| 29 | + - Generate synthetic ESIS-II images from simulations. |
| 30 | + - Add noise "correctly". |
30 | 31 |
|
31 | 32 | - Papers |
32 | 33 |
|
33 | | - - ESIS-I Instrument Paper (ASAP) |
34 | | - - CNN Inversion Paper |
35 | | - - ESIS-I Full Detector Inversion (big event analysis) |
36 | | - - ESIS-II pre-flight Science? |
37 | | - - ESIS-II mission/results |
38 | | - |
39 | | -Schedule: |
40 | | -We need and inversion pipeline that runs end to end by the end of April, experiment with it over late spring early summer, ready to write papers and show results at SPD. |
41 | | - |
42 | | -March: |
43 | | - |
44 | | -- Get a functioning distortion map for all channels at a single time |
45 | | -- Read distortion map in and form interpolated system |
46 | | -- Calculate weights arrays and transpose from interpolated system |
47 | | -- Define CTIS repo inversion skeleton and fundamental objects |
48 | | - |
49 | | -April: |
50 | | - |
51 | | -- Tune distortion mapping, check L3 difference images |
52 | | -- Program MART/RL method |
53 | | - |
54 | | -May/June: |
55 | | - |
56 | | -- Explore inversion parameter space |
57 | | -- Compare methods |
58 | | - |
59 | | -July: |
60 | | - |
61 | | -- Write inversion papers/prepare for SPD |
62 | | - |
63 | | - |
| 34 | + - ESIS-I Instrument Paper (ASAP) |
| 35 | + - CNN Inversion Paper |
| 36 | + - ESIS-I Full Detector Inversion (big event analysis) |
| 37 | + - ESIS-II pre-flight Science? |
| 38 | + - ESIS-II mission/results |
0 commit comments