Releases · SASfit/SASfit
- reduced the frequency of updating the progress bar as well as checking of pressing the interrupt button to maximal 10 times per second.
- added two additional form factors for ring polymers with excluded volume effects (RingPolymerBZ, RingPolymerBMB) and renamed the old one from FlexibleRingPolymer to RingPolymerCasassa
- bug fix in ring polymers, daisy-like ring and m-membered twisted ring.
- bug fix of GzSphere, returns wrong value for z>2R
- bug fix of GzgDAB for extrapolation to z->0
- added two more models for P'(Q) of thin cylindrical objects describing wormlike structure: "Koyama worm" and "freely joined chain of rods"
- Some form factors for polymers like Gaussian chains, star polymers and ring polymers have been moved to the plug-in area.
- As all peak functions are now available as plug-ins the obsolete old code has been deleted
- The Lorenz-Mie form factors for static light scattering are now made available as plug-in form factors
- SASfit now tries to avoids starting a new fit or simulation before the previous one is finished
- Successfully updated from gsl2.4 to gsl2.5 library.
- The new 2D/3D hard sphere structure factors are now available in the structure factor plug-in folder named "Hard Sphere 2D/3D"
- update to FFTW 3.3.8 worked after disabling threads however on Windows McAfee had to switched off during configure process.
- implemented some thermodynamic self-consistent hard sphere structure factors based on the rational function approximation method
- tried to avoid overflow of sinh-function in the model "Khodolenko-Worm"
- now between thin objects (local planar and local cylindrical approximation) and anisotropic shapes is distinguished.
- started to implement a 2D fit routine, which can fit data of a single detector position
- new plug-in folder for azimuthal data
- new plug-in folder for deformed or sheared objects
- new form factors for partly aligned cylinders and ellipsoids following a Maier-Saupe, Onsager, Heaviside, Gauss, or a Boltzmann orientation distribution
- new factor of a reptating chain after a deformation step
- new plug-in structure factors for 2-dimensional fluids of monodisperse hard spheres, "2D hard disks (Rosenfeld)" and "2D hard disks (Guo)"
- The form factors "Ellipsoid i" "Ellipsoid ii" "EllipsoidalCoreShell", and "triaxEllShell1" have been disabled and replaced by a series of plug-in functions. Those replacements make use of a faster routine for multidimensional integrals (pcubature). Because of this an optional size distribution became part of the supplied plug-in form factors.
- bug fix of unit conversion via clipboard
- added a plug-in for a random flight structure factor
- bug fix in the algorithm for calculating the resolution parameter in case of averaging neighboring data points
- plug-in implementation of several variants of helices
- plug-in for rectangular parallelepipeds have been extended so that it can have a size distribution of one, two or all three axis. The multiple integration is done by using the pcubature code from Steven G. Johnson
- multi dimensional integration package "cubature" from Steven G. Johnson becomes generally available also in plug-ins (https://github.com/stevengj/cubature)
- changed scaling of SESANS correlation functions by factor 1/(2pi)^2
- unit conversion for reading ASCII files corrected and extended
- update to FFTW 3.3.6-pl2 did not work, went back to 3.3.5
- update to gsl 2.4
- bug fix: resolution bar was not plotted properly
- Uncertainties on x and y axis are now plotted symmetrically from x+-Delta_x and y+-Delta_y.
Before they were plotted from x+-0.5Delta_x and y+-0.5Delta_y.
- moved all peak functions into the plugin area.
- Added the possibility to apply after the summation of all scattering contributions a final operator.
At the moment between three operator can be chosen:
- Unit operator which does not do anything with the data
- Taking the Hankel transform 2pi*H[I(Q)] for converting SAS model into a projected correlation function G(delta),
which defines the SESANS signal
- Applying the operator from 2) and taking the exponetial function of it exp(2pi*H[I(Q)]) to get SESANS signal
- For the extension to SESANS also an option for reading SESANS data has been supplied.
The format has been defined by TU Delft and has the default extension *.ses.
- added a model for of a self-fine random density distribution (gDAB) both under non-particular
structures as well as under SESANS as a correlation function.
- added a SESANS correlation function for a generalized Gaussian chain for Flory exponents between nu ϵ (0,1/2)
- updated to FFTW 3.3.5
- updated to sundials 2.7.0
- updated to gsl 2.3
- supplying some alternative integration routines for integrating over the size distribution
- bug fix of sasfit_integrate routine, which caused a crash when called from plugin functions
- first minimal routine for reading simple ascii data from ALV-5000 with single correlation function for DLS-analysis
- adding the PLHNC and RMSA closure to the OZ solver
- fixed a bug in the scaling parameter of the Teubner-Strey model
- bugfix in routine for reading data in BerSANS format
MacOS version requires X11 like xquartz
- implementation of another cumulant formula for DLS
- Bug-fix in the unit conversion routine
- implementation of GMRES, Bi-CGStab, TFQMR and Andersen Accelarion
for solving efficiently the Ornstein Zernicke fixpoint problem
- bug fix: since version 0.94.4 the new interrupt option suppressed
a proper error reporting to the GUI for undefined input values
- removed some old structure factors,
which were rarely used and theoretically not up to date
- a first version of a plugin for ordered mesoscopic and nano structures.
The plugin is providing a part of the structure factors available
in the software package "scatter" from S. Förster.
- plugin of a radial profile for a sphere resulting in a Porod law both below and above Q^-4
- improved analysis of confidence intervals of fit parameters:
- clickable correlation coefficients in the matrix highlight associated
pair of fit parameters
- covariance matrix elements are highlighted according to their correlation
- highlighting and selection of correlated params improved
- clickable correlation coefficients in the matrix highlight associated
- storing intensity and size distribution in batch processing routine
- changed the width of the error bar, assuming that the supplied error is FWHM
- changed the internal procedure for plotting error bars
- resolution parameter can now also be plotted
- changed the order of plotting the fit results
in "integral structural parameters"
- interruption of batch processing or series analysis implemented
- a new interrupt button can stop now the intensity evaluation of fitting
procedure after each q-value. Before the whole scattering curve needed to be
- progress bar has been added to the GUI
- bug fix in plotting error bars of multiple data sets.
- implementation of another method for thinning out oversampled data sets. The
new methods is performing an averaging of data points depending on a
user-defined maximum allowed q-smearing and a user-defined maximum distance
in intensity in units of the error bar of the data points, i.e. an averaging is
only performed, if the intensities look similar with n-times the intensity
- bug fix in GUI if one wants to forget old data and load a new data sets.
- new binding for zoom option, which works better for a mouse with one button
(plot zoom by ctrl+left-mouse btn)
- removed key binding for scattering contributions
- Added several iteration schemes for finding the fixpoint of the OZ equation.
- Added a tab for the total correlation function h(r)
- Added also some root finding algorithms for solving OZ equation.
- resolution parameter from file was not scaled during the change of units for Q
- bug fix in calculation of xi for sq for wormlike structures (PS1 model)
- wrong label for contrast in Teubner-Strey model
- bug fix in SPHERE+R^-a_Nagg form factor in assignement of core volume
- data export format set to scientific notation in the form x.yyyE?zz only