Fix Chabrier

[keflavich]

The Chabrier distributions were unclearly specified and used a set of constants I couldn't dig out of the papers, so I changed them to use constants I could find.

If anyone finds the coefficients I was missing, let me know.

[keflavich]

My old numbers actually came from McKee & Offner

[segasai]

Just out of curiosity, how different is the default new Chabrier from the previous one ? (I have been using the chabrier2005 myself). I clearly incorrectly referred to non-existing eqn in Chabrier2005. The sigma number for it (.55) is the one in given in eqn 2 of Chabrier2005. But the center was 0.2 for some reason rather than 0.25 given in the paper...

[keflavich]

good question; pretty sure the difference is minimal but i think it would be good to quantify.

mckee & offner quoted chabrier 2005 for the numbers, so if there is an error in our work, it's propagating theirs. but i think the distrs agree within uncertainty

On Thu, Mar 17, 2022 at 14:19 Sergey Koposov @.***> wrote:

Just out of curiosity, how different is the default new Chabrier from the previous one ? (I have been using the chabrier2005 myself). I clearly incorrectly referred to non-existing eqn in Chabrier2005. The sigma number for it (.55) is the one in given in eqn 2 of Chabrier2005. But the center was 0.2 for some reason rather than 0.25 given in the paper...

— Reply to this email directly, view it on GitHub https://github.com/keflavich/imf/pull/30#issuecomment-1071180611, or unsubscribe https://github.com/notifications/unsubscribe-auth/AABDCY3APFRQIM4F25KVOQDVANZTHANCNFSM5Q7KNVAA . You are receiving this because you modified the open/close state.Message ID: @.***>

-- Adam

[keflavich]

also we could keep the chabrier2005 ref around in the code to avoid breaking things downstream. thoughts?

On Thu, Mar 17, 2022 at 14:25 Adam @.***> wrote:

good question; pretty sure the difference is minimal but i think it would be good to quantify.

mckee & offner quoted chabrier 2005 for the numbers, so if there is an error in our work, it's propagating theirs. but i think the distrs agree within uncertainty

On Thu, Mar 17, 2022 at 14:19 Sergey Koposov @.***> wrote:

Just out of curiosity, how different is the default new Chabrier from the previous one ? (I have been using the chabrier2005 myself). I clearly incorrectly referred to non-existing eqn in Chabrier2005. The sigma number for it (.55) is the one in given in eqn 2 of Chabrier2005. But the center was 0.2 for some reason rather than 0.25 given in the paper...

— Reply to this email directly, view it on GitHub https://github.com/keflavich/imf/pull/30#issuecomment-1071180611, or unsubscribe https://github.com/notifications/unsubscribe-auth/AABDCY3APFRQIM4F25KVOQDVANZTHANCNFSM5Q7KNVAA . You are receiving this because you modified the open/close state.Message ID: @.***>

-- Adam

-- Adam

[segasai]

I think the options are

• keep chabrier2005 as an alias to the chabrier version you put in today
• preserve old chabrier2005 and put a big warning/deprecate

I'd say if you can make a plot comparing the new chabrier and old chab2005 and they are very similar, I'd go with the first option. Otherwise if the difference is noticeable I'd go with the second option.