Changes

Page history
Update General Review (post GW190814) authored by Reed Essick's avatar Reed Essick
Show whitespace changes
Inline Side-by-side
General-Review-(post-GW190814).md
View page @ 5bed15c7
......@@ -23,9 +23,9 @@ This review wiki covers the changes to the code based in order to incorporate po
* [ ] Monte-Carlo point-estimates and error estimates derived in [LIGO-P2000216](https://dcc.ligo.org/LIGO-P2000216) / [arXiv:2007.01372](https://arxiv.org/abs/2007.01372) are correct
* *Note!* Reed has since found smally typos in the error estimates published in these works (associated with covariance between "f" and "g", etc).
* **ACTION ITEM** (for Reed): update derivations to show correct error estimation and to include formulas for Bayes factor.
* [ ] Monte-Carlo algorithms are correctly implemented within calls to [utils.prob_odds](https://git.ligo.org/reed.essick/mmax-model-selection/-/blob/master/mmax_model_selection/utils.py#L335)
* [ ] Monte-Carlo algorithms are correctly implemented within calls to [utils.prob_odds](https://git.ligo.org/reed.essick/mmax-model-selection/-/blob/master/mmax_model_selection/utils.py#L434) (or [utils.post_prior_bayes](https://git.ligo.org/reed.essick/mmax-model-selection/-/blob/master/mmax_model_selection/utils.py#L476) as appropriate).
* [ ] population priors are correctly implemented within calls to [utils.population_weights](https://git.ligo.org/reed.essick/mmax-model-selection/-/blob/master/mmax_model_selection/utils.py#L107)
* [ ] reweighting of single-event posterior samples is by population priors is correctly done within the [mmax-model-selection executable](https://git.ligo.org/reed.essick/mmax-model-selection/-/blob/master/bin/mmax-model-selection#L138)
* [ ] reweighting of single-event posterior samples is by population priors is correctly done within the [mmax-model-selection executable](https://git.ligo.org/reed.essick/mmax-model-selection/-/blob/master/bin/mmax-model-selection#L153)
---
......@@ -64,8 +64,9 @@ Beyond parsing the command-line options, `mmax-model-selection` is relatively st
* this is accomplished through a delegation to [utils.population_weights](https://git.ligo.org/reed.essick/mmax-model-selection/-/blob/master/mmax_model_selection/utils.py#L107), which is little more than a set of nested conditionals that set up population parameters before evaluating the ratio of prior probabilities (between the target prior and the one used to generate the event's posterior samples) for the
* masses: [single-object mass distribution](https://git.ligo.org/reed.essick/mmax-model-selection/-/blob/master/mmax_model_selection/utils.py#L86) and [pairing probability](https://git.ligo.org/reed.essick/mmax-model-selection/-/blob/master/mmax_model_selection/utils.py#L97)
* distance: [uniform in euclidean volume](https://git.ligo.org/reed.essick/mmax-model-selection/-/blob/master/mmax_model_selection/utils.py#L100). *Note* we do not currently incorporate cosmological models to get the distribution in comoving volume.
* Compute the posterior odds and the posterior probability, along with error estimates through a delegation to [utils.prob_odds](https://git.ligo.org/reed.essick/mmax-model-selection/-/blob/master/mmax_model_selection/utils.py#L335). Because the Odds ratio and posterior odds are estimated using many of the same calculations (see Appendix of [LIGO-P2000216](https://dcc.ligo.org/LIGO-P2000216) / [arXiv:2007.01372](https://arxiv.org/abs/2007.01372), these are estimated within the same function call).
* Compute the posterior odds and the posterior probability, along with error estimates through a delegation to [utils.prob_odds](https://git.ligo.org/reed.essick/mmax-model-selection/-/blob/master/mmax_model_selection/utils.py#L434). Because the Odds ratio and posterior odds are estimated using many of the same calculations (see Appendix of [LIGO-P2000216](https://dcc.ligo.org/LIGO-P2000216) / [arXiv:2007.01372](https://arxiv.org/abs/2007.01372), these are estimated within the same function call).
* this computes the expected value and standard deviation through nested Monte-Carlo integrals over weighed samples.
* note, if prior samples are also provided, the code will compute the prior odds and Bayes factor in addition to the posterior odds.
The full help-string is reproduced below
......@@ -73,7 +74,13 @@ The full help-string is reproduced below
$ mmax-model-selection -h
usage: mmax-model-selection [-h] [--mass-column MASS_COLUMN]
[--m1-column M1_COLUMN] [--m2-column M2_COLUMN]
[--d-column D_COLUMN] [--mmax-column MMAX_COLUMN]
[--d-column D_COLUMN]
[--event-prior-samples EVENT_PRIOR_SAMPLES]
[--prior-mass-column PRIOR_MASS_COLUMN]
[--prior-m1-column PRIOR_M1_COLUMN]
[--prior-m2-column PRIOR_M2_COLUMN]
[--prior-d-column PRIOR_D_COLUMN]
[--mmax-column MMAX_COLUMN]
[--mmax-weight-column MMAX_WEIGHT_COLUMN]
[--mmax-weight-is-log]
[--mmax-range MMAX_RANGE MMAX_RANGE]
......@@ -82,7 +89,7 @@ usage: mmax-model-selection [-h] [--mass-column MASS_COLUMN]
[--mc-range MC_RANGE MC_RANGE]
[--d-range D_RANGE D_RANGE] [-p POPULATION]
[-P POPULATION_PARAMETER POPULATION_PARAMETER]
[-v]
[-v] [-V]
event_samples mmax_samples
a simple script to estimate bayes factors for a NS vs a BH based on m2
......@@ -99,6 +106,14 @@ event posterior arguments:
--m2-column M2_COLUMN
only used when pruning posterior samples
--d-column D_COLUMN only used when pruning posterior samples
--event-prior-samples EVENT_PRIOR_SAMPLES
--prior-mass-column PRIOR_MASS_COLUMN
--prior-m1-column PRIOR_M1_COLUMN
only used when pruning posterior samples
--prior-m2-column PRIOR_M2_COLUMN
only used when pruning posterior samples
--prior-d-column PRIOR_D_COLUMN
only used when pruning posterior samples
Mmax posterior arguments:
mmax_samples
......@@ -121,6 +136,7 @@ prior arguments:
verbosity arguments:
-v, --verbose
-V, --Verbose
```
### Other Executables
......
......