search

annayqho / TheCannon

a data-driven method for determining stellar parameters and abundances from stellar spectra
MIT License
39 stars 16 forks source link

where is the fit happening? #30

Closed davidwhogg closed 9 years ago

davidwhogg commented 9 years ago

In preparation for going over to a Gaussian Process (which will be a subsequent issue), what I want to know is: What are all the locations in the code that need to be changed if (say) we changed the code from a second-order fit to a third-order fit. Each of these locations should be marked with a comment that says # specialized to second-order model. Then I can just search for those and look at how we will go over to a Gaussian Process.

annayqho commented 9 years ago

OK, I think there's only two places (and actually those two places should probably be turned into one place). I added that comment to them. The directory is AnnaCannon/Code/Version1.3/ and the relevant scripts are: cannon/cannon1_train_model.py cannon/cannon2_infer_labels.py

davidwhogg commented 9 years ago

can you point me to line numbers where the form of the model matters most or is chosen?

also, why are there directories with Version numbers in them? That's what git does for you -- keep versions!

annayqho commented 9 years ago

line 130 in cannon1_train_model line 26 in cannon2_infer_labels

(and yes, sorry! I will stop making new versions from now on, I've been clinging to a bad pre-Github habit.)