SasView is now distributed through github. The official release site can be found at: https://github.com/SasView/sasview/releases
The latest release files are copied here for convenience of those using the "check for updates" feature in 3.0
- New in Version 3.1.0
- The documentation/help has had a complete overhaul including:
- A completely new presentation interface (Sphinx).
- Proof reading!
- Updating for latest features.
- A Help (or sometimes ?) button has been added to every panel, and some
sub panels if appropriate, linking to the appropriate section in the
- The model help has been split so that the Details button now brings up
a very short pop-up giving the equation being used while HELP goes to
the section in the full documentation describing the model.
- Extensive help has also been added for the new optimizer engine (see
below) including rules of thumb on how and when to choose a given
optimizer and what the parameters do.
- The optimizer engine has been completely replaced. The new optimizer
still defaults to the standard Levenberg-Marquardt algorithm. However 4
other optimizers are now also available. Each starts with a set of default
parameters which can be tuned. The DREAM optimizer takes the longest but
is the most powerful and yields rich information including full parameter
correlation and uncertainty plots. A results panel has been added to
- The five new optimizers are:
- A Levenberg-Marquardt optimizer
- A Quasi-Newton BFGS optimizer
- A Nelder-Mead Simplex optimizer
- A Differential Evolution optimizer
- A Monte Carlo optimizer (DREAM)
- New models were added:
- MicelleSphCoreModel (currently residing in the Uncategorized category)
- Existing models were updated:
- LamellarPS (bug in polydispersity integration fixed)
- Infrastructure to allow SESANS data to be fit with models was added. This
will become available in a future release but can currently be used from
the command line with some caveats.
- A number of bugs were fixed including a thread crashing issue and an
incorrect slit smearing resolution calculation.
- Implemented much more robust error logging to enable much easier
debugging in general but particularly the debugging of issues reported by
- A number of infrastructure tasks under the hood to enhance maintainability
- Upgrade from Wx 2.8 to Wx 3.0.2 which allows several new features but
required significant additional rework as well.
- Fully implemented Sphinx to the build process to produce both better
user documentation and developer documentation.
- Restructuring of the code base to more unified nomenclature and structure
so that the source installation tree more closely matches the installer
- Code cleanup (an ongoing task) .
- Migration of the repository to github simplifying contributions from
non-project personnel through pull requests.
- New in Version 3.0.0
- The GUI look and feel has been refactored to be more familiar for
Windows users by using MDI frames. Graph windows are also now free-
- Five new models have been added: PringlesModel, CoreShellEllipsoidXTModel,
RectangularPrismModel, RectangularHollowPrismModel and
- The data loader now supports ILL DAT data files and reads the full meta
information from canSAS file formats.
- Redefined convention for specifying angular parameters for anisotropic
- A number of minor features have been added such as permitting a log
distribution of points when using a model to simulate data, and the
addition of a Kratky plot option to the linear plots.
- A number of bugs have also been fixed.
- Save Project and Save Analysis now work more reliably.
- BETA: Magnetic contrast supporting full polarization analysis has been
implemented for some spherical and cylindrical models.
- BETA: Two new tools have been added:
- A generic scattering calculator which takes an atomic, magnetic or
SLD distribution in space and generates the appropriate 2D
scattering pattern. In some cases the orientationally averaged
(powder) 1D scattering can also be computed. Supported formats
include: SLD or text, PDB, and OMF magnetic moment distribution
- An image viewer/converter for data in image format; this reads in
an image file and will attempt to convert the image pixels to
data. Supported formats include: TIFF, TIF, PNG, BMP, JPG.
- New in Version 2.2.1
- Minor patch to support CanSAS XML v1.1 file format
- Added DataInfo for data in the DataExplorer and plots
- Added Maximize/Restore button in the title bar of the graphs
- Added a hide button in the toolbar of the graph panel
- The 'x' button now deletes a graph
- Edit SUM Model from the menubar can now generate and save more than one sum model
- Reports can now be saved in pdf format on WIN and MAC
- Made significant improvements to the batch/grid panel and fixed several bugs
- Fixed a number of other minor bugs
- New in Version 2.2.0
- Application name changed to SasView
- New fully customizable Category Manager added for better management of
increasing number of models
- Improved the Grid Window functionality in the batch fitting mode
- Added a simpler Graph/Plot modification interface
- Added a new 'Data Operation' tool for addition, subtraction, multiplication,
division, of two data sets.
- The 'Sum Model' editor was extended and renamed 'Summation and Multiplication'
- Added more plot symbols options for 1d plots
- Added improved trapping of compiling errors to the 'New model editor'
- Added some intelligent outputs (e.g., Rg, background, or rod diameter
depending on the choice of axis scale of the plot) to the linear fits
- Added more models
- Feature set from previous versions
- Perspectives Available
- Invariant calculator: Calculates the invariant, volume fraction, and
specific surface area.
- P(r) inversion calculator: Indirect Fourier transformation method.
- Fitting: the tool used for modeling and fitting 1D and 2D data to
analytical model functions
- Tools: provides a number of useful supplementary tools such as SLD
- Includes a large number of model functions, both form factors and structure factors.
- Support P(Q)*S(Q) for form factors that flag they can be so multiplied.
- Supports Gaussian, lognormal, Shulz, rectangular and custom distribution
functions for models that need to include polydispersity or for orientational
distributions if appropriate.
- Anisotropic shapes and magnetic moment modeling in 2D allow for a non-uniform
distribution of orientations of a given axis leading to modeling and fitting
capabilities of non azimuthaly symmetric data.
- User can choose to weight fits or not. If using weights, the user can choose
the error bar on each point if provided in the file, the square root
of the intensity or the intensity itself.
- Instrumental resolution smearing of model or fits is provided with several
options: read the resolution/point fromt he file. Input a pinhole resolution
or a slit resolution.
- Users can define the Qrange (Qmin and Qmax) for both 1D and 2D data for
fitting and modeling, but not graphically. The range can be reset to the
defaults (limits of q in data set for a fit) with the reset button.
- A mask can be applied to 2D calculation and fitting.
- Normalized residual plots are provided with every fit.
- Model function help available through detail button or from the fitting panel.
- Simultaneous/(advanced)constrained fitting allows for fitting a single
data set or several different sets simultaneously with the application
of advanced constraints relating fit parameters to functions of other
parameters (including from a different set). For example thickness of
shell = sin(30) times the length.
- Models that are the sum of two other models can be easily generated through the
SUM Model menubar item.
- New Python models can be added on the fly by creating an appropriate Python
file in the model plugin directory. Two tools are provided to help:
An easy to use custom model editor allows the quick generation of new Python
models by supplying only the parameters and their default value (box 1)
and the mathematical function of the model (box 2) and generating the
necessary *.py file. A separate advanced model editor provides a full Python
file editor. Either way once saved the model becomes immediately available
to the application.
- A batch fitting capability allows for the analysis of a series of data sets to
a single model and provides the results in a tabular form suitable for saving
or plotting the evolution of the fit parameters with error bars (from within
- A scattering length density calculator,including some X-ray information
- A density to vol. fraction converter is provided
- In application access to a Python shell/editor (PyCrust) is provided
- An instrument resolution calculator, including possible gravitational and
TOF effects is provided
- A slit size calculator optimized for Anton Paar Saxess is provided.
- A kiessig fringe thickness calculator is provided
- Plots and plot management
- A 3D graphing option (for 2d data/results) is provided with the view
controlled by the mouse
- 2D plots are shown with an intensity color bar. 2D Color map can be user
- Supports output of plot to a variety of graphic formats. Supported formats
include: png, eps, emf, jpg/jpeg, pdf, ps, tif/tiff, rawRGBbitmap(raw, rgba),
and scalable vector graphic (svg/svgz)
- Supports ouput of data in plot (1 or 2D) to limited data formats
- Multiple data sets can be loaded into a single graph for viewing (but a fit
plot can currently only have a single plot).
- Extensive context sensitive plot/fitting/manipulation options are available
through a right mouse click pop-up menu on plots.
- Data management
- Supports 2 + column 1D ASCII data, NIST 1D and 2D data, and canSAS data
via plug-in mechanism which can easily allow other readers as appropriate.
- 2D data is expected in Q space but for historical reasons accepts the
NIST 2D raw pixel format and will do conversion internally.
- The full data and metadata available to SasView is viewable in ASCII via
right clicking on a data set and choosing Data Info in the DataExplorer
or on the plots
- Supports loading a single file, multiple files, or a whole folder
- An optional Data Explorer is provided (default) which simplifies managing,
plotting, deleting, or setup for computation. Most functions however do
not require access to the explorer/manager and can be accessed through
right click menus and the toolbar. The data explorer can be re-started
from the menu bar.
- Data manipulation
- Support various 2D averaging methods : Circular, sectors, annular,
boxsum, boxQx and boxQy.
- A 2D data maks editor is provided
- 2D mask can be applied to the circular averaging.
- Miscellaneous features
- limited reports can be generated in pdf format
- Provides multiprocessor support(Windows only)
- Limited startup customization currently includes default startup
data folder and choice of default starting with data manager
- Limited support for saving(opening) a SasView project or a SasView analysis
(subproject) is provided.
- SasView can be launched and loaded with a file of interesty by double-clicking
on that file (recognized extension)
- A data file or data folder can be passed to SasView when launched from
the command line.
- Limited bookmarking capability to later recall the results of a fit calculation
- Extensive help is provided through context sensitive mouse roll-over,
information bar (at the bottom of the panel), the console menu, and
access to the help files in several different ways.
2- Downloading and Installing
*** Note 1: Much more information is available at www.sasview.org under links.
Look in the 'For Developers' section and particularly the wiki at
*** Note 2: If you have EXE or ZIP SasView installer, you won't need any of
the following. However it is highly recommended that any
previous versions be un-installed prior to installing the
2.1- System Requirements
- Python version >= 2.5 and < 3.0 should be running on the system
- We currently use Python 2.7
2.2- Installing from source
- Get the code from GitHub at https://github.com/SasView/sasview.git
for this release version use: https://github.com/SasView/sasview/releases
- run 'python setup.py install' under the 'sasview-x.x.x' folder
- run 'python sasview.py' under the 'sasview' folder.
- The following modules are required (version numbers are what are used
in the windows release build):
- Common Packages
- reportlab 3.1.44
- lxml 184.108.40.206 (MAC 220.127.116.11)
- PIL 1.1.7
- xhtml2pdf 3.0.33 (MAC = not installed on build server)
- unittest-xml-reporting 1.12.0 (MAC 1.10.0)
- matplotlib Version Installed: 1.4.3 (MAC 1.1.1)
- bumps Version Installed: 0.7.5.6
- scipy Version Installed: 0.16.0b2 (MAC 0.11.0)
- periodictable Version Installed: 1.4.1
- setuptools Version Installed: 7.0 (MAC 12.0.5)
- sphinx Version Installed: 1.3.1 (MAC 1.3b2)
- pyparsing Version Installed: 2.0.3
- numpy Version Installed: 1.9.2 (MAC 1.6.2)
- html5lib Version Installed: 0.99999
- wx Version Installed: 18.104.22.168
- Windows Specific Packages
- pywin 219
- py2exe 0.6.9
- comtypes 1.1.1
- MinGW w/ gcc version 4.6.1 (WIN)
- vcredist_x86.exe (version 9.0.21022.8 -microsoft visual C 2008
- Innosetup (WIN - isetup 5.4.2-unicode) - used to create distributable
*** Note: Windows build dependencies can be set up using anaconda. Instructions
can be found at http://trac.sasview.org/wiki/AnacondaSetup
- MAC Specifc Packages
- py2app 0.7.1
3- Known Issues
3.1- All systems:
- The documentation window may take a few seconds to load the first time
it is called. Also, an internet connection is required before
equations will render properly. Until then they will show in their
original TeX format.
- If the documentation window remains stubbornly blank, try installing a
different browser and set that as your default browser. Issues have
been noted with Internet Explorer 11.
- Check for Updates may fail (with the status bar message ' Cannot
connect to the application server') if your internet connection uses
a proxy server. Tested resolutions for this are described on the
- The copy and paste functions (^C, ^V) in the batch mode results grid
require two clicks: one to select the cell and a second to select the
contents of the cell.
- The tutorial has not yet been updated and is somewhat out of date
- Very old computers may struggle to run the 3.x and later releases
- Polydispersity on multiple parameters included in a simultaneous/
constrained fit will likely not be correct
- Constrained/simultaneous fit page does not have a stop button
- Constrained/simultaneous fit do not accept min/max limits
- Save project does not store the state of all the windows
- Loading projects can be very slow
- Save Project only works once a data set has been associated with
a model. Error is reported on status bar.
- There is a numerical precision problem with the multishell model when
the iner radius gets large enough (ticket #288)
- The angular distribution angles are not clearly defined and may in
some cases lead to incorrect calculations(ticket #332)
- If installed to same directory as old version without first removing
the old version, the old desktop icon will remain but point to the
new exe version. Likewise all the start menu folders and items will
have the old name even though pointing to the new version. Usually
safest to uninstall old version prior to installing new version anyway.
- Application normally starts up hidden. Click icon in Dock to view/use
- Multiprocessing does not currently work on MAC OS
- Not well tested
4- SasView website
- www.sasview.org. This main project site is the gateway to all
information about the sasview project. It includes information
about the project, a FAQ page and links to all developer and user
information, tools and resources.
5- Frequently Asked Questions
6- Installer download website
- Latest release Version
- Latest developer builds