MgII and CIV as alternative models for LLS search

[jibanCat]

Adding MgII and CIV to the LLS search

• gp_find_lls : the main search file

For the CIV and MgII metals, the search range should be the same as LLS in the Lambda_obs space.

So, for each metal, need to do:

• set_parameters: We can create a new set_parameters file, but the search range z_civ samples should be (1 + z_hi) * lambda_hi / lambda_mgii - 1
• model_priors: no such thing - maybe an order of magnitude estimate from DLAs or search for CDDF
• samples : this is just uniform, can be pre-computed and load from .h5 files

[jibanCat]

Model priors for metals:

My DLA CDDF is

Mg II CDDF is (https://iopscience.iop.org/article/10.3847/1538-4357/abbb34/pdf)

The fit was performed only on column densities above the region of partial completeness (log N(Mg II)>12.4 cm−2) so as to not skew the slope. We obtained δ=1.45±0.01.

CIV CDDF is (https://arxiv.org/pdf/1511.01094.pdf)

Kim 2013 (https://arxiv.org/pdf/1302.6622.pdf)

The black dotted line shows the linear regression to filled circles at log NC iv = [12.25, 15.5] : log dnC iv/(dNC ivdX) = (11.41±1.61)+(−1.85±0.13)×log NC iv. The solid line is the power law fit at lo

[jibanCat]

Models priors:

p(MgII | zQSO) ~ 10^12 p(HI | zQSO)
p(CIV | zQSO) ~ 10^10 p(HI | zQSO)

Parameter priors

p(log N) \propto 10^ log(dn / dXdN)

[jibanCat]

Lyman limit systems:

My suggest: cut off at 21, -2 for DLAs, and connecting at each breaking point

[jibanCat]