This is a part of the code that I used in my thesis, to determine the absolute magnitude of small amplitude red giants. Details regarding the procedure can be found here:
https://obswww.unige.ch/~palavers/Thesis/thesis_col.pdf
(Chapter 3, and in particular 3.6 and 3.8).
Briefly, this particular and simplified example uses only OGLE III LMC small amplitude RGs. Basically, it just tests if the method works, since the difference between the true and infered mag should be zero. Code does the following:
- Picks a "test" star.
- Searches for all the remaining stars that have the same combo of top three periods that the test star has (within +-2.5% in logP).
- Peak of the mag distribution of the selected stars (i.e. those with the same periods) is considered as the indicator of the "true" absolute magnitude.
To run the code, RG.py needs to be executed.
Black points in the left panel represent the entire sample. Yellow crosses represent the position of the test star in PL diagram. Transparent (R,G,B) bins indicate +-2.5% around the three periods of the test star (logP).
Black points in the middle panel represent only the stars with the "three-period combo" matching that of the test star. Yellow dashed line is the true mag of the test star, cyan dashed line is the inferred mag (ideally they should be overlapping since the code is "reconstructing" LMC).
Right panel shows the difference between true and inferred mag (blue patch between yellow and cyan lines), and the normalized smoothed histogram (black) and KDE (red) of the magnitude distribution of the stars from the middle plot.