Duty #64098
Updated by Kardum Leonora 26 days ago
Agenda:
- preconditioner for the polychromatic MapMaking
- projection operators
--- The convergence might be due to pixelisation effect which is the next subtask.
- TOD creation problem: pixellisation effect
--- Tom sees better performance with higher nside and with larger convolution. These arguments partly support the idea the convergence issue is coming from the pixelisation effect. Implementing a interpolation function in pysimulators might be difficult so JC suggest a handwritten method of building TOD. Future tasks are:
1. test on truly uniform polarization Q+U
2. Convergence vs Nside (>512)
3. Build TOD by hand for 1 timesample
Once the issue is solved, it should be well described in a note that Tom can use as a chapter for the PhD.
- uniforming reading dictionaries
--- Steve is working on get_dict across qubicsoft
- Qnoise
- randomness in code
--- We need to check if the seed and seed_noise are properly initialized at the beginning of the TOD building