@@ -104,14 +104,14 @@ After some bugs were identified, a rerun of the millilensing run has yielded the
...
@@ -104,14 +104,14 @@ After some bugs were identified, a rerun of the millilensing run has yielded the
After additional bugs were squashed, it was identified that the testing set may be simply insufficiently strong to allow detection. A more "obvious" test was constructed by setting luminosity_distance = 1200 and dl1 = 410 (inverse of before) and has been run yielding the following results
After additional bugs were squashed, it was identified that the testing set may be simply insufficiently strong to allow detection. A more "obvious" test was constructed by setting luminosity_distance = 1200 and dl1 = 410 (inverse of before) and has been run yielding the following results
| Injection Model | Analysis Model | Run Directory | Corner Plot | Sign Off (Suvodip) | Sign Off (Harsh) |
| Injection Model | Analysis Model | Run Directory | Corner Plot | Sign Off (Suvodip) | Sign Off (Harsh) |
Parameter recovery is significantly improved from before, with support for the true values of all parameters in the posteriors. Individual phases are not distinctly recovered, but results would indicate that the relative phase difference would be recovered correctly.
Parameter recovery is significantly improved from before, with support for the true values of all parameters in the posteriors. Individual phases are not distinctly recovered, but results would indicate that the relative phase difference would be recovered correctly.
This leads to log Bayes factors favouring the millilensing model of 362.75 over the microlensing model and 435.72 over the unlensed model, also meaning a favouring of the microlensing model of 72.97 which would again indicate the validity of using the point mass microlensing model as the initial sweep.
This leads to log Bayes factors favouring the millilensing model of 362.75 over the microlensing model and 435.72 over the unlensed model, also meaning a favouring of the microlensing model of 72.97 which would again indicate the validity of using the point mass microlensing model as the initial sweep.
### Candidate O4 Sampling Review Tests
### Candidate O4 Sampling Review Tests
...
@@ -130,7 +130,7 @@ These runs use the configuration from the microlensing true run with the candida
...
@@ -130,7 +130,7 @@ These runs use the configuration from the microlensing true run with the candida
Sign offs:
Sign offs:
* [ ] Harsh
* [ ] Harsh
* [ :white_check_mark:] Suvodip
* [ ] \[:white_check_mark:\] Suvodip
#### Bilby MCMC
#### Bilby MCMC
...
@@ -148,9 +148,9 @@ Sign offs:
...
@@ -148,9 +148,9 @@ Sign offs:
| Lens Model | Run Status | Run Directory | Amplification Plot | Sign Off (Harsh) | Sign Off (Suvodip) |
| Lens Model | Run Status | Run Directory | Amplification Plot | Sign Off (Harsh) | Sign Off (Suvodip) |
