Line-Profile Analysis Software (LIPRAS), is a graphical user interface for least squares fitting of Bragg peaks in diffraction data. For any region of the inputted data, user can choose which profile functions to apply to the fit, constrain profile functions, and view the resulting fit in terms of the profile functions chosen. A Bayesian inference analysis can be carried out on the resulting least squares result(s) to generate a full description of the errors for all profile parameters.
Authors: Giovanni Esteves, Klarissa Ramos, Chris Fancher, and Jacob Jones
• Quickly extract relevant peak information: position, full width at half maximum (FWHM), and intensity
• Conduct Bayesian inference on least squares results using a Markov Chain Monte Carlo algorithm (need Statistics and Machine Learning Toolbox)
• Analyzes files with a different number of data points and/or X-values. However, check fitting range before attempting.
• Customize the background fit by either treating it separately (Polynomial or Spline) or including it in the least squares routine (Polynomial only)
• Choose from 5 different peak-shape functions: Gaussian, Lorentzian, Pseudo-Voigt, Pearson VII, and Asymmetric Pearson VII.
• Peak-shape functions can be constrained in terms of intensity, peak position, FWHM, and mixing coefficient
• Automatically calculate Cu-Kalpha2 peaks when working with laboratory X-ray data
• Fit up to 20 peaks in the current profile region
• For multiple diffraction patterns, results from the previous fit can be used as the subsequent starting parameters for next fit
• Visualize results with a plot of the resulting peak fit and residual plot, allowing you to see what peaks make up the overall fit
• Resulting coefficients values can be viewed with file number to quickly view trends in data
• Parameters files can be written and used to recreate fits and details what fit parameters and profile shape functions were used
• Accepts the following file types: .xy, [.ras, .acs] (Rigaku), .xls, .xlsx, .fxye, .xrdml (Panalytical), .chi, .csv (Windows Only)
LIPRAS is currently updated through GitHub
Web Page: https://github.com/SneakySnail/LIPRAS
Requires MATLAB 2016b, Curve Fitting Toolbox, and GUI Layout Toolbox to run. The Statistics and Machine Learning Toolbox is required for Bayesian analysis.
If you use LIPRAS for your research, please cite it (choose one):
1. Giovanni Esteves, Klarissa Ramos, Chris M. Fancher, and Jacob L. Jones. LIPRAS: Line-Profile Analysis Software. (2017). DOI: 10.13140/RG.2.2.29970.25282/3
2. Giovanni Esteves, Klarissa Ramos, Chris M. Fancher, and Jacob L. Jones. LIPRAS: Line-Profile Analysis Software. (2017). https://github.com/SneakySnail/LIPRAS
Great tool ! Fitting several diff. patterns at a time is very nice.
Thank you AJ, I really appreciate it. I will definitely check out your submission.
Great GUI the least squares fitting works great! I just put out a GUI toolbox check it out. I think you may like it.
1. The 'Update' button, after conducting a fit, was not applying results of LS fit as starting parameters on next LS (i.e. hitting fit data) on the first click
1. Read in Rigaku ASC and RAS files
1. Analyze files with a different number of data points in least squares and Bayesian component. Check fitting range before attempting.
1. Fixed an issue that caused the Bayesian of analysis of multiple files to use the SP from the first file and applied it to all files.
Added multipliers to SD and UB/LB bounds
Bug fixes to Bayesian GUI. Added ability to plot Sigma^2 trace plot.
Bug Fixes, Right-click now delete Bkg points during selection, Introduced alpha version of Bayesian analysis component, which requires Statistics and Machine Learning Toolbox
The ‘Check for Updates’ now displays current version and the available version if an update is recommended. Added a try/catch statement to .csv, .xlsx, .xls files for handling 2 or 3 column data (twotheta, intensity), (twotheta, intensity, error).
Added an if statement for cases when measured intensity was 0, which would cause the weight of that value to be 'Inf' in MATLAB.
Significant improvement to background refinement when 'Refine Bkg' is on
Small changes, 'Default' weight options for CHI files were incorrect. This was a typo, although could be forced by specifying 1/obs. Found out GOF or (reduced chi-square) is sse/dfe which are values found in the Fmodel files.
See Update Notes, Highlight:
Highlights, See update notes for details
Needs Curve Fitting Toolbox to run
SourceForge now has stand-alone version
Added GitHub webpage
Added Stand-Alone version link for those that want to run LIPRAS without MATLAB
Inspired by: GUI Layout Toolbox
Create scripts with code, output, and formatted text in a single executable document.