SasView 4.1 RC2
Pre-releaseThis point release brings a series of new features and improvements,
and a host of bug fixes. Of particular note are:
Correlation Function Analysis (Corfunc)
This performs a correlation function analysis of one-dimensional SAXS/SANS data,
or generates a model-independent volume fraction profile from the SANS from an
adsorbed polymer/surfactant layer.
A correlation function may be interpreted in terms of an imaginary rod moving
through the structure of the material. It is the probability that a rod of
length R moving through the material has equal electron/neutron scattering
length density at either end. Hence a frequently occurring spacing within a
structure manifests itself as a peak.
A volume fraction profile describes how the density of polymer
segments/surfactant molecules varies with distance from an (assumed locally flat)
interface.
Fitting of SESANS Data
Data from Spin-Echo SANS measurements can now be loaded and fitted. The data will
be plotted against the correct axes and models will automatically perform a Hankel
transform in order to calculate SESANS from a SANS model.
Documentation
The documentation has undergone significant checking and updating.
Acknowledgement and citation
If you publish work making use of SasView 4.0, please include the following text:
This work benefited from the use of the SasView application, originally developed under NSF Award DMR-0520547. SasView also contains code developed with funding from the EU Horizon 2020 programme under the SINE2020 project Grant No 654000.
SasView can be cited as :
M. Doucet, et al. SasView Version 4.1, Zenodo, http://doi.org/10.5281/zenodo.438138
New Features
- Correlation function (corfunc) analysis of 1D SAS data added from CCP13
- File converter tool for multi-file single column data sets
- SESANS data loading and direct fitting using the Hankel transformation
- Saving and loading of simultaneous and constrained fits now supported
- Save states from SasView v3.x.y now loaded using sasmodel model names
- Saving and loading of projects with 2D fits now supported
- Loading a project removes all existing data, fits, and plots
- Structure factor and form factor can be plotted independently
- OpenCL is disabled by default and can be enabled through a fit menu
- Data and theory fields are now independently expandable
Bug Fixes
- Fixes #667: Models computed multiple times on parameters changes
- Fixes #673: Custom models override built in models of same name
- Fixes #678: Hard crash when running complex models on GPU
- Fixes $774: Old style plugin models unloadable
- Fixes #789: stacked disk scale doesn't match cylinder model
- Fixes #792: core_shell_fractal uses wrong effective radius
- Fixes #800: Plot range reset on plot redraws
- Fixes #811 and #825: 2D smearing broken
- Fixes #815: Integer model parameter handling
- Fixes #824: Cannot apply sector averaging when no detector data present
- Fixes #830: Cansas HDF5 reader fully compliant with NXCanSAS v1.0 format
- Fixes #835: Fractal model breaks with negative Q values
- Fixes #843: Multilayer vesicle does not define effective radius
- Fixes #858: Hayter MSA S(Q) returns errors
- Numerous grammatical and contexual errors in documention
Current Known Issues
Listed here are known issues that users should be aware of. A full list of known bugs and feature requests is available at http://trac.sasview.org/report/3
Critical
- No currently known critical issues
Other Significant
- The way that orientation is defined is being refactored to address long standing issues and comments. In release 4.1 however only models with symmetry (e.g. a=b) have been converted to the new definitions. The rest (a <> b <> c - e.g. parellelepiped) maintain the same definition as before and will be converted in 4.2. Note that orientational distribution also makes much more sense in the new framework. The documentation should indicate which definition is being used for a given model.
- The infrastructure currently handles internal conversion of old style models so that user created models in previous versions should continue to work for now. At some point in the future such support will go away. Everyone is encouraged to convert to the new structure which should be relatively straight forward and provides a number of benefits.
- In that vein, the distributed models and those generated by the new plugin model editor are in the new format, however those generated by sum|multiply models are the old style sum|multiply models. This should also disappear in the near future.
- The on the fly discovery of plugin models and changes thereto behave inconsistently. If a change to a plugin model does not seem to register, the Load Plugin Models (under fitting -> Plugin Model Operations) can be used. However, after calling Load Plugin Models, the active plugin will no longer be loaded (even though the GUI looks like it is) unless it is a sum|multiply model which works properly. All others will need to be recalled from the model dropdown menu to reload the model into the calculation engine. While it might be annoying it does not appear to prevent SasView from working..
Model Issues
Corrections will be posted to Marketplace as available (and fixed in following release)
- Core shell parrallelipiped fails to calculate the long (C side) shell. If set non-zero this leads to an incorrect volume normalization and thus to an incorrect absolute scaling. This has apparently always been true. If thick_C is set to zero the model behaves correctly.
- Triaxial ellipsoid does not reproduce the limiting cases of oblate or prolate ellipsoid.
Bug Reporting
Please support the project by reporting bugs that you find to [email protected]
All the known bugs/feature requests can be found at: http://trac.sasview.org/report/3
Regular developer builds are also available from https://jenkins.esss.dk/sasview/ if you wish to test the very latest (most likely unstable) versions of SasView.
[This release makes use of sasmodels v0.96]