Geronimo Estellar Chen
Well, I do want to classify this two types of supernovae solely based on spectra. Based on Daniel Muthukrishna's DASH, I and Stoner are interested in recognizing type IIP and IIL SNe solely via their spectra. Currently, I choose a Functional Primary Component Analysis algorithm FPCA to straighten out the problem, however, it is not quick enough. If it really works, I will fulfill the documentation and upload the program --but who knows...
I have managed to classify Type IIb and IIP SNe from their functional major components, both SMV and neuro-network reaches a acceptable accuracy. The re-constructed spectra looks faithful to the original data.
More-Data, More-Data, MORE------DATA!!! I have tried to download as much as I can.
Maybe I need a better pipeline between python and R?
I need to properly categorized the data I have acquired.
If there is truely something substantial, I will publish it.
Well, Microsoft Excel is not an ideal programme for handling with csv files, for it may add something implicitly.
I am trying to fix this problem by removing these steps, because FPCA works in the data without re-sampling. That will make the whole process more detailed, but may cause extra calculation.
Many thanks to Prof. Liu Xuewen and predecessor Stoner for teaching me python and R language. Moreover, many thanks to Thomas Russell for his generator. In the last, I really appreciate Prof. Avishay Gal-Yam and Prof. Lifan Wang for the guidance.