====== Telecon Notes Nov. 29 2018 ======

Attendance: Graca, Mathieu, Shaul  \\
Regrets:  \\

Notes: Shaul \\

=== Updates ===
  * __Mathieu__ ran 20 simulations of 90.92 with r=0, 85% delensing, 36% sky coverage. The results are in the logbook. 
    * We see significant residuals above lensing at \ell<20. Discussing what's going on. 
    * SH asks: is the excess power at low \ell the result of small sky coverage, or of the actual foreground residual. How does the pseudo cl code deal with small sky coverage?  Actions: 
      * Mathieu will plot the input CMB for the same mask 36% mask
      * Mathieu will use a much larger 98%, and much smaller 2.5% mask. 
      * Mathieu will show the masks. 
    * We note that if we believe the results, the foreground residual is white below \ell=200, and happens to peak above the delensed at \ell<10. Is that the limit for GNILC cleaning? 
    * Mathieu says that the code has not been optimized yet.
  * __Graca__ ran Soumen's r=0 maps. She gets r=0, but the dispersion is 10^{-7}. Far too low. 
    * Suggestion: take a step back; validate code on pure CMB simulations with no foregrounds, but with noise. 
      * without foregrounds, recover r=0 with full lensing, and with 85% delensing. 
      * without foregrounds, recover r=1e-3 with 85% delensing with small and larger noise.