cmfgen2tardis not working

[Knights-Templars]

Code Sample, a copy-pastable example if possible

Problem description

from tardis.atomic import AtomData
ModuleNotFoundError: No module named 'tardis.atomic'

While trying to convert a cmfgen file to tardis, shows the output above. Any help would be greatly appreciated

[wkerzendorf]

could you please show us how you want to do that and post also the files - so we can try to follow

[Knights-Templars]

Thankyou for your reply.

So, I want to model the spectra of a SN Ia which I am currently working on. I am trying to generate tardis readable file from a cmfgen file
PDDEL1n_SN_HYDRO_DATA_1.000d.txt

This is from a paper by Dessart, 2014.
I am running cmfgen2tardis /cmfgen-file /tardis-model as is described in the manual. I am getting this error

   from tardis.atomic import AtomData
ModuleNotFoundError: No module named 'tardis.atomic'

I also tried importing atomic from jupyter-notebook. It is not working there also.

[epassaro]

Seems AtomData has been moved:

from tardis.io.atom_data import AtomData

Also in the script tardis/scripts/cmfgen2tardis.py there's a line:

AtomData.from_hdf5()

but currently AtomData has the to_hdf method instead. Also, you need to point to the atomic data file.

AtomData.from_hdf('/home/user/tardis-data/kurucz_cd23_chianti_H_He.h5')

After doing that I ran the script by doing:

cmfgen2tardis PDDEL1n_SN_HYDRO_DATA_1.000d.txt .

and got the following error:

ValueError: could not convert string to float: '100*0.00000D0'

because Python does not understand Fortran D type. I recommend change that values directly in the .txt file.

This script should be updated soon.

[Knights-Templars]

Many thanks. I did the import as was mentioned by @epassaro , and also removed the string part '100*0.00000D0' to 100.00.
Now a new error comes as follows on the same text file that was provided earlier.

abundances_df.loc[element_symbol] = abundances

ValueError: cannot set a row with mismatched columns

Is there a change in the output format of cmfgen or the code ?

[epassaro]

I understand that '100*0.00000D0' should be 0.0 (no abundance).

Then the ValueError happens because abundances for elements different than '100*0.00000D0' have the same length. Then what we need (I'm not really sure) is an array of zeros of the same length.

I wrote on line 80:

                if len(abundances) == 1:
                    abundances_df.loc[element_symbol] = [0.]*len(density)   # A big row of 0s
                else:
                    abundances_df.loc[element_symbol] = abundances

Then the code fails again because the function def get_atomic_number returns index -1 by default (line 14 and 69) and there's no element with Z=-1.

In line 70 added:

                if atomic_no == -1:
                    break

and got the output file . But I'm not sure if this approach makes sense, I never worked with this script.

@wkerzendorf knows more about this.

[wkerzendorf]

I'll have a look hopefully soon