Skip to content

mishioo/tesliper

Repository files navigation

Build status Documentation Status Coverage Status PyPi version PyPI pyversions Code style: black License DOI

tesliper

tesliper: Theoretical Spectroscopist's Little Helper is a program for batch processing of Gaussian output files, focused on calculations of vibrational, electronic, and scattering spectra from Gaussian-calculated quantum properties of molecule's conformers.

It allows to easily exclude conformers that are not suitable for further analysis: erroneous, not optimized, of higher energy or lower contribution than a user-given threshold. It also implements an RMSD sieve, enabling one to filter out similar structures. It lets you calculate theoretical IR, VCD, UV, ECD, Raman, and ROA spectra for each conformer or as a population-weighted average and export obtained spectral data in one of supported file formats: .txt, .csv, or .xlsx. Finally, if allows to easily setup a next calculations step with batch export to .gjf Gaussian input files.

tesliper is written in Python 3.6 and makes use of some additional third party packages (see below or requirements.txt). It may be used as a package or as a stand-alone application with dedicated GUI.

Getting Started

You can use tesliper from python or as standalone application with dedicated graphical user interface. See below for details.

Requirements

Python 3.6+
numpy
openpyxl
matplotlib (optional, for GUI)

tesliper uses tkinter to deliver the graphical interface. It is included in most Python distributions, but please be aware, that some might miss it. You will need to install it manually in such case.

Installing to your Python distribution

tesliper is available on PyPI, you can install it to your python distribution by simply running:

$ python -m pip install tesliper

or

$ python -m pip install tesliper[gui]

if you would like to be able to use a graphical interface.

A standalone application

This option is currently available only for Windows users. To get your copy of tesliper, simply download a Tesliper.exe file from the latest relase. This file is a standalone application, no installation is required.

How to use

Full documentation is available online: https://tesliper.readthedocs.io/.

Primer

Conventions that are important to note:

  • tesliper stores multiple data entries of various types for each conformer. To prevent confusion with Python's data type and with data itself, tesliper refers to specific kinds of data as "genres". Genres in code are represented by specific strings, used as identifiers. To learn about data genres known to tesliper, see Available data genres.
  • tesliper identifies conformers using a stem of an extracted file (i.e. its filename without extension). When files with identical names are extracted in course of subsequent Tesliper.extract calls or in recursive extraction using Tesliper.extract(recursive=True), they are treated as data for one conformer. This enables to join data from subsequent calculations steps, e.g. geometry optimization, vibrational spectra simulation, and electronic spectra simulation. Please note that if specific data genre is available from more than one calculation job, only recently extracted values will be stored.
  • tesliper was designed to deal with multiple conformers of single molecule and may not work properly when used to process data concerning different molecules (i.e. having different number of atoms, different number of degrees of freedom, etc.). If you want to use it for such purpose anyway, you may set Tesliper.conformers.allow_data_inconsistency to True. tesliper will then stop complaining and try to do its best.

In Python scripts

Tesliper class is the main access point to tesliper's functionality. It allows you to extract data from specified files, provides a proxy to the trimming functionality, gives access to data in form of specialized arrays, enables you to calculate and average desired spectra, and provides an easy way to export data.

Most basic use might look like this:

from tesliper import Tesliper
tslr = Tesliper()
tslr.extract()
tslr.calculate_spectra()
tslr.average_spectra()
tslr.export_averaged()

This extracts data from files in the current working directory, calculates available spectra using standard parameters, averages them using available energy values, and exports to current working directory in .txt format.

You can customize this process by specifying call parameters for used methods and modifying Tesliper's configuration attributes:

  • to change source directory or location of exported files instantiate Tesliper object with input_dir and output_dir parameters specified, respectively. You can also set appropriate attributes on the instance directly;
  • to extract only selected files in input_dir use wanted_files init parameter. It should be given an iterable of filenames you want to parse. Again, you can also directly set an identically named attribute;
  • use Tesliper.conformers.trim... methods to easily filter out conformers you wish to ignore in further analysis;
  • to change parameters used for calculation of spectra, modify appropriate entries of parameters attribute;
  • use other export methods to export more data and specify fmt parameter in method's call to export to other file formats.
tslr = Tesliper(input_dir="./myjob/optimization/", output_dir="./myjob/output/")
tslr.wanted_files = ["one", "two", "three"]  # only files with these names
tslr.extract()  # use tslr.input_dir as source
tslr.extract(path="./myjob/vcd_sim/")  # use other input_dir
tslr.conformers.trim_not_optimized()  # filtering out unwanted conformers
tslr.parameters["vcd"].update({"start": 500, "stop": 2500, "width": 2})
tslr.calculate_spectra(genres=["vcd"])  # we want only VCD spectrum
tslr.average_spectra()
tslr.export_averaged(mode="w")  # overwrite previously exported files
tslr.export_activities(fmt="csv")  # save activities for analysis elsewhere
tslr.output_dir = "./myjob/ecd_sim/"
tslr.export_job_file(  # prepare files for next step of calculations
    route="# td=(singlets,nstates=80) B3LYP/Def2TZVP"
)

A graphical interface

If you are using tesliper as a standalone application, simply double click on the Tesliper.exe file to start the application. To invoke it from the command line, just run tesliper-gui. GUI consists of three panels and a number of controls. The panels are: "Extracted data", "Energies list", and "Spectra view". First two offer a list of conformers read so far using "Chose files" and "Chose folder" buttons on the left. The last enables to preview calculated spectra.

screenshot

  • "Extracted data" panel shows an identifier of each conformer (a file name) and an overview of data extracted. Little checkboxes on the left of each conformer may be clicked to mark this conformer as "kept" or "not kept".
  • "Energies list" offers the same set of conformers and checkboxes, but with energies values listed for each conformer. The view may be changed using "Show" dropdown box in "Energies and Structure" section of controls, to present difference in energy between conformers or their percentage contribution in population.
  • "Spectra view" tab shows calculated spectra. It may be controlled using "Calculate spectra" section. After choosing a spectra genre to calculate you may control if it is simulated using lorentzian or gaussian fitting function, change peak width, spectra bounds, etc. You may view spectra for one conformer, all of them stacked, or averaged. You may also load an experimental spectrum (.txt or .csv format) for comparison.

screenshot

Once done with extracting files and tweaking parameters, export selected data to desired format or save the session for later using buttons in "Session control" section.

A detailed tutorial with screenshots is available in the documentation: https://tesliper.readthedocs.io/en/stable/gui.html.

License

This project is licensed with BSD 2-Clause license. See LICENSE.txt file for details.

Contributing to tesliper

Contributions are welcome! tesliper is a growing project and definitely has room for improvements.

Bugs and suggestions

Bug reports are of great value, if you encounter a problem please let me know by submitting a new issue. If you have a suggestion how tesliper can be improved, please let me know as well!

Participating in code

If you'd like to contribute to tesliper's codebase, that's even better! If there is a specific bug that you know how to fix or a feature you'd like to develop, please let me know via issues. To start coding, get your working copy of tesliper's source code by cloning the repository and then setup your environment by installing development dependencies (probably to a virtual environment):

python -m pip install .[dev]

Please remember to add/update relevant tests along with your code changes. Make sure the test suite passes by running

python -m pytest test

or, if you'd like to also generate the coverage report, you can invoke the pytest with the pytest-cov plugin (included in the "dev" installation), as shown below. The .coveragerc file includes plugin's configuration (i.e. to omit GUI code).

python -m pytest --cov=tesliper --cov-config=.coveragerc test

If you contribute code to tesliper, the test coverage badge will be updated by the CI service after your changes are merged.

Although mypy is not incorporated in tesliper's development (yet!), I believe that type hints greatly improve the experience of using the package. Please, add them to the new code you submit. If you're willing to supplement typing of the existing code, that would be very much welcome!

tesliper's codebase is formatted with black and isort, please use these tools before submitting your code contribution. To make it easier, you may use a pre-commit configuration available in this repository. To include it in your workflow, simply run pre-commit install in your copy's root directory. Note, that this configuration also sets up flake8 linter.

To get your change introduced to the codebase, please make a Pull Request to the fixes branch for quick bug fixes or to the dev branch for new features and bigger changes. If at a loss, do not hesitate to reach to me directly! :)

Roadmap

Ideas for possible future improvements to the software are listed below. Based on the feedback from the Community, I will decide, which ones are desired and worth working on.

New Functionality:

  • command line interface
  • support for Jaguar & other packages
  • option for using cclib for parsing
  • spectra comparison feature
  • velocity and length electronic spectra comparison
  • parsing conformational search files
  • GUI: manually choose spectra colour

UX Improvements:

  • supplement logging
  • auto finding optimal spectra range
  • GUI: export spectra as image
  • GUI: spectra colour by population
  • GUI: drag&drop support
  • GUI: display tooltips on hover

Acknowledgements

Many thanks to the scientists, who advised me on the domain-specific details and helped to test the software:

as well as to people, who reviewed the project: @alejandrogallo and @arepstein.

How to cite

I'm very happy to announce, that tesliper has been published in the Journal of Open Source Software! If you find tesliper useful in your research, please give it a credit by citing it:

  • Więcław, M. M., (2022). tesliper: a theoretical spectroscopist's little helper. Journal of Open Source Software, 7(72), 4164, DOI: 10.21105/joss.04164
@article{tesliper2022,
  doi = {10.21105/joss.04164},
  url = {https://doi.org/10.21105/joss.04164},
  year = {2022},
  publisher = {The Open Journal},
  volume = {7},
  number = {72},
  pages = {4164},
  author = {Michał M. Więcław},
  title = {tesliper: a theoretical spectroscopist's little helper},
  journal = {Journal of Open Source Software}
}