search

dgrin1 / axionCAMB

MIT License
19 stars 7 forks source link

Matter Power Spectrum Fault #11

Closed Saber-of-QFT closed 2 years ago

Saber-of-QFT commented 2 years ago

Hello, I met some weird problem when running axionCAMB.

I try to plot matter power spectrum with different axion mass or different axion fraction, to compare with Fig 2. in arXiv:1410.2896. I set parameter in params.ini as follow (for example, fixed f_ax = 1, and varying m_ax) : image

But I got some ridiculous result : image image

As I shown above, the power spectrum has an overall supression on all scales, where the large scale supression is unexpected.

But if I change the value of parameter 'omega_axion' as follow : image

I set the value of 'omega_axion' equal to 'omaxh2', which I think it is make no sense physically. But I obtain a result consistent with the Fig 2. in arXiv:1410.2896 : image image

This result is very puzzling, have you met this problem before? Or did I set some cosmological parameters in a wrong way?

dgrin1 commented 2 years ago

Hi there- is use physical units set to true or false? I think what is happening is that as you vary one parameter, the code is adjusting H_0 to hold something else fixed - this leads to (expected) changes in the growth function. That would explain why things act as expected when you vary the other parameter. You may have to switch the setting on physical_units flag and things should work then. Let me know what happens

Saber-of-QFT commented 2 years ago

Hi, in my original ini file, I set use_physical = F, because I want to use Omega_m, Omega_b, Omega_nu ... (and so no) as my input parameters.

I try to set use_physical = T, and using ombh2, omch2, omnuh2 and omaxh2 as input parameters, then the result is right now.

Does it mean if we want to include axion in our model, we have to set physical units to true ? Thank you !

dgrin1 commented 2 years ago

I wouldn't put it that way - I would just say you have to properly interpret the results. If you set use_physical=false, the code will be adjusting H_0 in ways that dramatically alter the growth function - the results should be correct for that particular cosmology, its just that this may not be the cosmology you want to actually compute.

Saber-of-QFT commented 2 years ago

Understood. I will carefully setting parameters in my model. Thank you so much for helping!