Skip to content

FinalAssessment.md

Latest commit

 

History

History
34 lines (23 loc) · 2.21 KB

FinalAssessment.md

File metadata and controls

34 lines (23 loc) · 2.21 KB
Raw

Back to Model selection

Final assessment

The Finalizer module

The final step in the modeling run is the classification of the microlensing event through comparison of models obtained for each category. This task is performed by a specific external module called Finalizer. This module can be launched by the corresponding function called Finalizer():

import RTModel
rtm = RTModel.RTModel('/event001')
rtm.Reader()
rtm.InitCond()
rtm.launch_fits('PS')
rtm.ModelSelector('PS')
...
rtm.Finalizer()

With this code, we first perform all steps as detailed in the previous sections and then we make the final assessment through Finalizer().

In the /event001 directory you will see that the following products appear:

  • A new subdirectory called FinalModels/ is created. This will contain the final proposed models for the microlensing event, which are just copies of those appearing in the directory Models/ that have passed all critieria.
  • A file nature.txt containing a summary of the chi square achieved in each model category, a final assessment, and the list of the competing models that have passed all criteria and copied to the directory FinalModels/. For these models, also the $\Delta \chi^2$ of the largest unmodeled feature is reported. This can be useful to assess the needfor further investigations.

The execution of Finalizer closes the modeling run. You may then proceed to plotting the best models.

Making the final assessment

The Finalizer module compares models of different categories according to Wilks' theorem, which states that the level at which a model with p additional parameters is preferred can be assessed by the chi square distribution with p degrees of freedom. More details can be read in the RTModel paper. Models that are not nested one within the other are compared with the softer threshold given by the 1-sigma level in the chi square distribution.

There are no available options for Finalizer() since an unsatisfied user may easily vet the models in the directory Models/ according to his/her own criteria.

Go to Astrophotometric fits