Describe the bug
when i run lobsterpy automatic-plot --allbonds
it reported
/data/software/anaconda/2020.07/envs/lobsterpy/lib/python3.8/site-packages/pymatgen/electronic_structure/cohp.py:164: UserWarning: This method has not been tested on many examples. Check the parameter limit, pls!
warnings.warn("This method has not been tested on many examples. Check the parameter limit, pls!")
The compound CrN has 1 symmetry-independent atoms(s) with relevant bonds: Cr1.
Cr1 has a single (CN=1) coordination environment. It has 1 Cr-Cr (mean ICOHP: -0.2 eV, antibonding interaction below EFermi) bonds.

what does it mean ? the Cr-Cr is anti bonding? but i cannot seen it in my picture. And it did not generate any output? How can I do with it.I asked before, but I didn't check the reply in time. I'm very sorry to ask again

A tutorial with the basic functions of lobsterpy might be useful for users that have never done a Lobster computation.

We need to update the installation process description in the readme.

Hi,
when I run lobsterpy automatic-plot ,it reported as followed,
/data/software/anaconda/2020.07/envs/lobsterpy/lib/python3.8/site-packages/pymatgen/electronic_structure/cohp.py:164: UserWarning: This method has not been tested on many examples. Check the parameter limit, pls!
warnings.warn("This method has not been tested on many examples. Check the parameter limit, pls!")
and it did not generate any results.How can solve the problem.thanks
zhou chao

There should be a command-line option to easily plot projected DOS from Lobster.

@naik-aakash, I think we can stop supporting Python 3.8.

As the tests are currently taking forever to run, we (@naik-aakash) should probably also look into speeding up the tests and linting. I had in mind to maybe also switch to a ruff linter.

better to use the snakecase naming convention consistently in the code

Todo

• A clear description of how orbital-wise automatic analysis works in our documentation, the tutorials and the paper
• Showcase the results for several examples of different complexity

Describe the bug
when plotting things with lobsterpy via lobterpy autoplot or lobsterpy plotdos, the following error is thrown:
OSError: '/home/certural/miniconda3/envs/atomate2/lib/python3.10/site-packages/lobsterpy/plotting/lobsterpy_base.mplstyle' not found in the style library and input is not a valid URL or path; see style.available for list of available styles

To Reproduce
e.g. lobsterpy plotdos

Expected behavior
Plotting nice plots

PS: this issue is more for documentation @naik-aakash is already working on a fix :)

Recently with the new and temporary students (that are also rather inexperienced with using cli etc.) I think a lot about user friendliness. So I thought maybe it is a good idea to have a default functionality for each lobsterpy feature e.g. lobsterpy plotdos will automatically do something like lobsterpy plotdos --elements [list of atoms from POSCAR] instead of throwing an error. The default should be chosen in a way to get most out of it with minimum complexity (e.g. in case of the DOS plotting elementwise DOS is more common than plotting orbitalwise DOS and also maybe easier to comprehend for beginners (I'm not sure about the last part, but when I learned this stuff a good while ago, this was true for me)).
Since I try to improve my programming skills in python (and C++) anyways by doing exercises on the weekend or so, I thought maybe I could implement it the next WE, if you (@JaGeo and @naik-aakash) are ok with this?

It was noted by @QuantumChemist , that createinputs alias set for cli does not work

Currently, lot of not necessary arguments show up for differnent possible actions. Could be cleanup a bit by regrouping the args

There are a few lines in the cli.py that are not yet covered by tests including

        if args.addtotaldos:
            dos_plotter.add_dos(dos=lobs_dos, label="Total DOS")

)
Maybe, we could address this.

Currently it takes long time for each run.

Atomate2 still supports 3.8 and our latest release does not leading to failing tests.

Is your feature request related to a problem? Please describe.
For large structures, there should be the option to skip the showing of the plots and only the save option should be used.

This has to be tested and also added to command line tools.

Describe the bug
Tests seem to fail. They worked for the pull request but failed afterwards, @naik-aakash .

Describe the bug
lobsterpy automatic-plot-ia --label-resolved
@naik-aakash , it raises an error and I am not sure where it comes from.

Add possibility to obtain orbital-resolved features in featurizer . Should be useful to get data for orbital interaction diagrams as well

• COXX fingerprints
• COXX center, width, skew, kurtosis

Describe the bug

Currently, saving figures from the automatic analysis does not work correctly. Only the last plot of the analysis will be saved.

We deprecated "get_anion_types" in pymatgen. We should therefore switch to the new command.

We need to update pymatgen, LobsterPy and atomate2 at some point in the future to allow for all new functionalities in Lobster 5.0.

Currently not all keys do not have snake_case convention: got over looked in paper

As lot of things are changed compared to our publication, maybe better to fix them now itself.

Also for example some keys like : charges and bandoverlaps could be renamed to be bit more descriptive as our larger schema already same keys with different data in it on top level

What do you think @JaGeo ?

It would be good to get our automatic analysis working for QE and Abinit outputs as well.

@davSchnieders, it would be great if you could provide a sample output. We could then make automatic analysis work for QE as well.

Describe the bug

I have to add a clearer error desciption when materials without cations/anions are investigated with this tool. Currently, the cation-anion mode will simply fail.

Lot of test files have been added,

update tests and clean up to reduce the size of repo

Hi @naik-aakash , we should probably also add the featurizers in the package 😃

This would allow to easily switch to use compressed files for the analysis with LobsterPy.

After merging all the PRs and reorganizing test files, update the old test durations files for optimized pytest runs

Allow that the user can provide their own basis functions for lobsterin generation - similar to the current pymatgen implementation.

Add all potential options for bonds that are also available in pymatgen:

e.g., only cation-cation bonds

Is your feature request related to a problem? Please describe.
I would like to implement an automatic analysis including orbital-resolved interactions. I will add this as an additional option so that people can also skip the orbital-resolved analysis.

I hope to accomplish this in the next 4 weeks.

Is your feature request related to a problem? Please describe.

• Implement the chemist's way of DOS plotting (DOS on x)
• add options to plot all orbitals per site
• add everything to the documentation

get POTCAR, INCAR, KPOINTS, lobsterin from pymatgen.io.vasp.LobsterSet automatically based on a POSCAR

Describe the bug
I think we would need to test these features more (e.g., test on compounds where we know what should come out)

This will allow to also automatially plot COOPs and COBIs but would require additional changes in pymatgen first.

Add something to inspect a lobsterout and other lobsteroutputs so that user's can be confident about their lobster results.

1. option to add all orbitals for a site to dos plot
2. option to plot invert axis dos

Add an option to the commandline to generate lobsterin and compute number of bands based on a POSCAR and POTCAR file. Doing this via pymatgen or within atomate is currently a bit labor-intensive.