Skip to content

Releases

pySED v2.4.1 - legacy 2.x

Choose a tag to compare

@Tingliangstu Tingliangstu released this 30 Oct 11:26

pySED 2.4.1: legacy 2.x maintenance release

✨ Highlights

This release preserves the existing 2.x workflow before the separate 3.0 upgrade.

⚡ Fixes since v2.4.0

  • Preserve explicitly supplied q-path fractions such as 7/15 and -17/31 through input parsing and commensurate q-point construction (#66).
  • Retain short-decimal normalization and support for programmatically supplied floating-point q paths.
  • Include regression tests for exact fractional paths and documentation/publication-list updates.

🧩 Workflow and Compatibility

  • The 2.x input interface, HDF5 workflow, numerical conventions, and MIT license are unchanged.
  • The v2.4.1 tag identifies this exact release; legacy/2.4 preserves the old-version branch.
  • Version 3.0 remains on codex/pysed-3.0-architecture. No 3.0 implementation has been merged into main or included here.

🚀 Installation

Install this fixed version from GitHub:

python -m pip install "git+https://github.com/Tingliangstu/pySED.git@v2.4.1"

Alternatively, download and install the attached wheel. The PyPI package named pysed is an unrelated text-processing utility; this release is published on GitHub, not PyPI.

📦 Downloads and Archives

  • GitHub's Source code archives contain the tagged repository, including examples and documentation sources.
  • The attached wheel and sdist contain the installable Python package.
  • pySED-2.4.1-manual-html.zip contains the frozen HTML manual; extract it and open index.html.
  • SHA256SUMS.txt provides SHA-256 checksums for the three attached distribution/manual files.

✅ Validation

Validated locally on Windows with Python 3.13:

  • All 5 legacy regression tests passed.
  • Strict Sphinx documentation build passed.
  • Wheel/sdist build and twine check passed.
  • Isolated wheel installation and all three CLI aliases passed.
  • Four synthetic HDF5 calculation cases passed: serial/parallel and total/partial SED agree, with the expected 6.25 THz peak.

🔎 Full changes: v2.4.0...v2.4.1

pySED v2.4.0

Choose a tag to compare

@Tingliangstu Tingliangstu released this 04 Jun 11:35

Changelog

v2.4.0 - 2026-06-04

✨ Highlights

  • pySED 2.4.0 improves commensurate q-path handling, fixes triclinic supercell box handling for LAMMPS-style workflows, and ships a much more complete online manual.

⚡ Core Fixes

  • Added support for fractional q_path coordinates such as 1/3 in input_SED.in.
  • Added tolerant handling of short decimal q-path approximations, so values close to small-denominator high-symmetry points, such as 0.33333, are treated as the intended commensurate fractions. This fixes q-path precision issues around commensurate-point detection (#41).
  • Fixed incorrect supercell box handling for LAMMPS-style restricted triclinic cells by deriving the full supercell matrix, converting it to the restricted representation, and applying the matching coordinate rotation for LAMMPS data output (#48).
  • Improved primitive-cell consistency checks when comparing the expected supercell box with the trajectory box.

🧩 Workflow and Compatibility

  • Fixed and clarified links in the MoS2 starting guide and example documentation.
  • Expanded guidance for partial SED workflows, including atom-type and direction-resolved examples linked from issue #39.
  • Expanded LO-TO splitting guidance with trajectory-dependence examples linked from issue #31.

🎨 Documentation and Examples

  • Reworked the ReadTheDocs manual structure with clearer navigation for introduction, requirements, installation, starting guide, tutorials, examples, input parameters, theory, tips, publications, reference, and troubleshooting.
  • Added a full input_SED.in parameter guide to the online manual, with dedicated pages for parameter syntax, defaults, examples, and related settings.
  • Added theory and practical tips pages to the documentation, including SED formulas, commensurate q-point construction notes, Lorentzian lifetime notes, LO-TO splitting guidance, and run-script workflow tips.
  • Expanded example documentation for CNT, graphene, MoS2, silicon, phonopy references, legacy LAMMPS workflows, and Jupyter tutorials.
  • Updated the root README to point users to the expanded online manual and parameter guide.
  • Updated the publication list with additional 2026 papers and corrected publication metadata.

2.3.0

Choose a tag to compare

@Tingliangstu Tingliangstu released this 01 Feb 07:07
8f35038

Overview

This release focuses on richer SED visualization controls, partial SED analysis, and improved structure/trajectory handling for better input compatibility and reproducibility.

✨ New
Partial SED output and plotting by atom type and Cartesian direction, enabling more granular analysis. (#39)

🎨 Visualization
New input tags to control SED colormap and colorbar range for clearer plots. (#29, #30)

⚡ Performance
Trajectory compression is faster, improving preprocessing for large datasets. (#40)

🧩 Compatibility & Fixes
Added support for reading .xyz structure files. (#28)
Fixed atomic order consistency between generated LAMMPS data and POSCAR. (#27)

2.2.0

Choose a tag to compare

@Tingliangstu Tingliangstu released this 14 Sep 15:58

🔔 pySED v2.2.0 Release Notes

We’re excited to announce the release of pySED v2.2.0, packed with enhancements, bug fixes, and improved usability.

Here's what's new:

🛠️ Bug Fixes & Input Handling

  • Resolved several reported issues to improve overall stability.
  • Enhanced user input validation for smoother and more informative interactions.
  • For a full list of closed issues, check the #21 #22 #24 #26.

📈 Improved k-point Generation & q-distance Calculation

  • Refined the logic for k-point generation, delivering more accurate and consistent results.
  • Introduced a new function to compute q-distance more reasonably, leading to better visualization and analysis.

📚 Updated Examples

  • Added a suite of new examples tailored to this version, making it easier to understand and apply pySED’s features.
  • Explore the examples here.

📖 Documentation Refresh

  • The user manual has been updated to reflect all changes and new functionalities introduced in v2.2.0.

v2.1.0

Choose a tag to compare

@Tingliangstu Tingliangstu released this 04 Aug 14:20
  1. Bug Fix: A major bug has been fixed where the phonon lifetimes were overestimated in previous versions due to a missing unit conversion. The issue has been resolved here: commit link.
    This fix does not affect the qualitative results from previous versions, and the updated lifetimes are now in line with prior literature.

  2. Improved Lorentzian Fit Initialization: The initialization parameters for Lorentzian fitting have been optimized for better performance.

  3. Minor Bug Fixes: Several small bugs have been addressed to improve the overall stability.

Full Changelog: v2.1.0...v2.1.0

v2.0.1

Choose a tag to compare

@Tingliangstu Tingliangstu released this 26 Dec 15:51
  1. Fixed bugs when reading files.
  2. Use more silky.

2.0.0

Choose a tag to compare

@Tingliangstu Tingliangstu released this 10 Dec 14:49
  1. Supports simultaneous fitting of multiple paths with ‘q_path_name’ and ‘q_path’.
  2. MD simulation supporting NPT with ‘rescale_prim = 1’.
  3. Add more use examples.
  4. Added the user manual.

pySED-0.3.0

Choose a tag to compare

@Tingliangstu Tingliangstu released this 29 Nov 08:23
  1. Add two flags for better phonon lifetime fitting (lorentz_fit_cutoff and modulate_factor).
  2. Add more tips in input_SED.in for phonon lifetime fitting.
  3. Add a silicon example for fitting the phonon lifetime of pristine silicon (see Silicon example). All the peaks fitted pretty well. But the SED method in 2010 always underestimates phonon lifetime, since fitting overestimates hwhm.

pySED-0.2.1

Choose a tag to compare

@Tingliangstu Tingliangstu released this 10 Oct 15:48
  1. Fix some bugs in v-0.2.0.

pySED-0.2.0

Choose a tag to compare

@Tingliangstu Tingliangstu released this 10 Oct 15:06
  1. Fix some bugs for plotting figures.
  2. When plot plot_cutoff_freq and plot_interval in [input_SED.in].(https://github.com/Tingliangstu/pySED/blob/main/example/Silicon/SED/input_SED.in) to help users adjust the frequency of SED images.
  3. When output q-points for view, change it to reduced-qpoints.
  4. Add a script plot_SED_figures.py to help plot nice image.