Wednesday, July 7, 2010

HR fitting for three parameters and new cluster

I have chosen two new stars in NGC 1528. They are both completely cluster members and around mass 3.1 solar masses. They have unidentified metalicities.

In the arena of chosen stars, I need to find metallicities for more of the clusters, or figure out how to calculate it. I also need to figure out how to calculate necessary exposure times. I want to conceptually work through the process of figuring it out instead of just looking it up, so that will take some thinking. I will continue to add new clusters to the list.

I also worked on extending the least squares fitting to cover distance, age, and reddening. Yesterday I managed to correct my distance fitting mechanism. What I had been doing previously was ignoring algebraic rules and I redid the code correctly and it worked very well. Today I devised a way to compile many age variables into one multidimensional array to use in the fitting process. This allowed me to successfully fit for both distance and age when keeping reddening fixed at the know value of 0.227 . Sadly when I extended it to cover all three variables, distance, age, and reddening, it didn't work. Age just went to the upper limit and reddening was off from the previously stated value significantly as well. Just looking at the plot showed it to be wrong as well.

I marginalized over the parameters in tau squared space, and the peak the code was returning for reddening was not the peak that the plot showed. I don't know what is going on. Age was just an increasing line, but somehow distance still worked quite well. I need to figure out why this it not working. I believe the outliers may somehow be throwing it off, so I need to work a process to account for them into my code.

This is getting to take quite some time to compute, with 11 ages, and 100 for each reddening and distance, with many computations at each combination. To speed things up and hopefully get a code that works, I intend to write a C++ code to fit HR diagrams. Sadly, I do not know how to compile and run the code on my computer since I learned on a fully set up linux computer. So I need to figure that out and write the code. I also need to figure out how to import the multidimensional age array into the C++ code.

No comments:

Post a Comment