exojax
exojax copied to clipboard
HajimeKawahara
🐈 Automatic differentiable spectrum modeling of exoplanets/brown dwarfs using JAX, compatible with NumPyro and JAXopt
Consider implementing the algorithm of [RotBroadInt](https://github.com/Adolfo1519/RotBroadInt) as a rotational spectral operator. https://arxiv.org/abs/2305.09693
I would like to try `kernex` to compute the convolution of the instruments/rotation, and maybe PreSolar convolution algorithmm. https://github.com/ASEM000/kernex
Note: I'd like to try flowMC https://github.com/kazewong/flowMC as a sampler of ExoJAX posterior.
I think something missing from [this notebook](https://github.com/HajimeKawahara/exojax/blob/master/documents/tutorials/Forward_modeling_for_Fe_I_lines_of_Kurucz.ipynb) is to divide `qt_K` determined in cell 5 by `adbK.QTref_284[mask]` before passing to `SijT` (or now `line_strength`) in the next cell, as the...
`searchsorted` strategy in `amclouds.compute_cloud_base_pressure_index` does not work. I need to think about the implementation of clouds. ```sh rg_layer, MMRc = amp_nh3.calc_ammodel( File "/home/kawahara/exojax/src/exojax/atm/atmphys.py", line 111, in calc_ammodel pressure_base = pressures[ibase]...
I will remove `install_radis_develop()` from `setup.py` after the new release of radis https://github.com/radis/radis/pull/691 (currently in radis/develop).
I noticed `in moldb.py`, now there are only atomic databases. We should reorganize Adb classes.
To support numpy 2, at least, we need to modify `radis`? The following bug was fixed when using `numpy 1.26` instead of `numpy 2.0.1` ```sh shirochan:~/exojax/tests/unittests/spec/api(radis0.15)>python3.11 molinfo_test.py An NVIDIA GPU...
There has been no emulate mdb for HITRAN yet. So, the unit test cannot be applied for HITRAN for api.py.
Metadata
Owner
Metadata
🐈 Automatic differentiable spectrum modeling of exoplanets/brown dwarfs using JAX, compatible with NumPyro and JAXopt