Skip to content

Changes

Page history
Update S191216ap authored by Sayantani Datta's avatar Sayantani Datta
Show whitespace changes
Inline Side-by-side
O3b-analysis/S191216ap.md
View page @ 633df66b
......@@ -29,11 +29,13 @@ Doubts :
Discussions:
Archishman: low q- high q regions may be coming from one or few faulty chains:
Archishman: low q- high q regions may be coming from one or few faulty chains.
>> No particular chain seem to be standing out much as can be seen [here](https://ldas-jobs.ligo.caltech.edu/~sayantani.datta/LSC/lalinference/O3/o3_b_tgr/sim/SummaryPages/S191216ap/C01/Prod10/run01/chains/home.html).
> No particular chain seem to be standing out much as can be seen [here](https://ldas-jobs.ligo.caltech.edu/~sayantani.datta/LSC/lalinference/O3/o3_b_tgr/sim/SummaryPages/S191216ap/C01/Prod10/run01/chains/home.html).
[GWTC-3.0 PE. channel on mattermost](https://chat.ligo.org/ligo/pl/du9w99kz1bbotg4m9jy1goi8aa): The regions where there is some trouble sampling might not be showing up for [pbilby](https://ldas-jobs.ligo.caltech.edu/~pe.o3/LVC/o3b-catalog/S191216a/Prod8/results/html/Prod8_Prod8.html) PE runs due to large number of samples. Our bbh run is somewhat consistent with a [preliminary run done with pv2](https://ldas-jobs.ligo.caltech.edu/~hsin-yu.chen/LVC/projects/O3/S191216ap/8s/html/EXP0_EXP0_mass_ratio.html), [Anjali's RIFT run](https://ldas-jobs.ligo.caltech.edu/~anjali.yelikar/O3_PEsumm_RIFT/O3b/S191216a/SEOBNRv4PHM_manual_noprocessing/html/RIFT-SEOBNRv4PHM_RIFT-SEOBNRv4PHM.html). Ajit Mehta did some checks with IMRXPHM comparing LI-mcmc, LI-nest and bilby runs. LI shows a peak at low q that bilby does not, see [here](https://hypatia.aei.mpg.de/~ajit.mehta/LVC/O3b_CBC_TGR/S191216ap/GR/peSummary/4samples/html/Comparison_mass_ratio.html).
[GWTC-3.0 PE. channel on mattermost](https://chat.ligo.org/ligo/pl/du9w99kz1bbotg4m9jy1goi8aa):
> The regions where there is some trouble sampling might not be showing up for [pbilby](https://ldas-jobs.ligo.caltech.edu/~pe.o3/LVC/o3b-catalog/S191216a/Prod8/results/html/Prod8_Prod8.html) PE runs due to large number of samples. Our bbh run is somewhat consistent with a [preliminary run done with pv2](https://ldas-jobs.ligo.caltech.edu/~hsin-yu.chen/LVC/projects/O3/S191216ap/8s/html/EXP0_EXP0_mass_ratio.html), [Anjali's RIFT run](https://ldas-jobs.ligo.caltech.edu/~anjali.yelikar/O3_PEsumm_RIFT/O3b/S191216a/SEOBNRv4PHM_manual_noprocessing/html/RIFT-SEOBNRv4PHM_RIFT-SEOBNRv4PHM.html). Ajit Mehta did some checks with IMRXPHM comparing LI-mcmc, LI-nest and bilby runs. LI shows a peak at low q that bilby does not, see [here](https://hypatia.aei.mpg.de/~ajit.mehta/LVC/O3b_CBC_TGR/S191216ap/GR/peSummary/4samples/html/Comparison_mass_ratio.html).
As there is a possibility of not so well sampled regions for our runs, I have launched runs with maxmcmc=10000.
......