- users - SasView Mailing Lists

[Sasview-users] subscribe
by Andrew Jackson 22 Aug '17

22 Aug '17

________________________________________________________________________ Andrew Jackson Instrument Scientist - Small Angle Scattering European Spallation Source, P.O Box 176, SE-221 00 Lund, Sweden Adjunct Associate Professor (Ajungerad Lektor) Physical Chemistry, Lund University, P.O. Box 124, SE-221 00, Lund, Sweden Phone: +46 46 888 3015 Mobile: +46 72 179 2015 E-mail: andrew.jackson@...631...<mailto:andrew.jackson@...631...> www.esss.se<http://www.esss.se/>

1 0

[Sasview-users] how to run sasview on Ubuntu 14.04 or 16.04 LTS 64 bit
by Martin Schmiele 19 May '17

19 May '17

Dear Users, has anybody succeeded to run sasview 4.1.0 on Ubuntu/Debian? I downloaded, built and installed the current source code versions for sasview + sasmodels from github. I can start the program, load data and specify fitting models (e.g. sphere model and using a polysphere example data set coming with the installation), but as soon as I start the fitting the window freezes and the whole application quickly terminates with a segmentation fault. I have the same problem on Ubuntu 14.04 (wx 2.8) and 16.04 (generally higher version numbers for the dependencies, wx 3.0). Thank you in advance, Martin p.s. I installed the dependencies listed on http://trac.sasview.org/wiki/LinuxBuildDependencies via Ubuntus package manager and pip. In case of the Ubuntu 16.04 installation, using "python check_packages.py" the following installed dependencies are listed: Checking Required Package Versions.... Common Packages reportlab Version Installed: 3.3.0 lxml Version Installed: 3.5.0.0 PIL Version Installed: 1.1.7 xhtml2pdf Version Installed: 3.0.33 pyopencl Version Installed: 2015.1 unittest-xml-reporting Version Installed: 1.5.0 pylint Installed (Unknown version) matplotlib Version Installed: 1.5.1 bumps Version Installed: 0.7.6 scipy Version Installed: 0.17.0 periodictable Version Installed: 1.4.1 setuptools Version Installed: 20.7.0 h5py Version Installed: 2.6.0 sphinx Version Installed: 1.5.5 pyparsing Version Installed: 1.5.7 numpy Version Installed: 1.11.0 html5lib Version Installed: 0.999 wx Version Installed: 3.0.2.0

2 1

Re: [Sasview-users] Log(I), log(q) fitting or I,q fitting?
by Andrew Jackson 18 Jan '17

18 Jan '17

Also, if you want to view the data as I vs Q, you can right click on the plot and choose “Change Scale” then select x vs y (rather than log(x) vs log(y) Andrew ______________________________________ Andrew Jackson Instrument Scientist - Small Angle Scattering European Spallation Source, P.O Box 176, SE-221 00 Lund, Sweden Adjunct Associate Professor (Adjungerad Lektor) Physical Chemistry, Lund University, P.O. Box 124, SE-221 00, Lund, Sweden Phone: +46 46 888 3015 Mobile: +46 72 179 2015 E-mail: a<mailto:andrew.jackson@...631...>ndrew.jackson@...631... www.esss.se From: "irenal@...634..." <irenal@...634...> Reply-To: Mailing list for users of SasView <sasview-users(a)lists.sourceforge.net> Date: Wednesday 18 January 2017 at 15:08 To: "users@...2..." <users@...2...> Subject: [Sasview-users] Log(I), log(q) fitting or I,q fitting? Hello, I'm a new user of this great program and I would like to ask how the fitting is done: Is it applied to the log-log curve (meaning that the natural logarithm of the intensity and the scattering vector are fitted- log(I), log(q) fitting) or directly on the intensity and the scattering vector (I,q fitting) ? I'm interesting in I,q fitting, and not in log(I), log(q) fitting. Thank you, Irena Levin

1 0

Re: [Sasview-users] Log(I), log(q) fitting or I,q fitting?
by Andrew Jackson 18 Jan '17

18 Jan '17

Hi Fitting is always done as I vs Q, regardless of the plotting display. Best ANdrew ______________________________________ Andrew Jackson Instrument Scientist - Small Angle Scattering European Spallation Source, P.O Box 176, SE-221 00 Lund, Sweden Adjunct Associate Professor (Adjungerad Lektor) Physical Chemistry, Lund University, P.O. Box 124, SE-221 00, Lund, Sweden Phone: +46 46 888 3015 Mobile: +46 72 179 2015 E-mail: a<mailto:andrew.jackson@...631...>ndrew.jackson@...631... www.esss.se From: "irenal@...634..." <irenal@...634...> Reply-To: Mailing list for users of SasView <sasview-users(a)lists.sourceforge.net> Date: Wednesday 18 January 2017 at 15:08 To: "users@...2..." <users@...2...> Subject: [Sasview-users] Log(I), log(q) fitting or I,q fitting? Hello, I'm a new user of this great program and I would like to ask how the fitting is done: Is it applied to the log-log curve (meaning that the natural logarithm of the intensity and the scattering vector are fitted- log(I), log(q) fitting) or directly on the intensity and the scattering vector (I,q fitting) ? I'm interesting in I,q fitting, and not in log(I), log(q) fitting. Thank you, Irena Levin

1 0

[Sasview-users] Log(I), log(q) fitting or I,q fitting?
by irenal＠...634... 18 Jan '17

18 Jan '17

Hello, I'm a new user of this great program and I would like to ask how the fitting is done: Is it applied to the log-log curve (meaning that the natural logarithm of the intensity and the scattering vector are fitted- log(I), log(q) fitting) or directly on the intensity and the scattering vector (I,q fitting) ? I'm interesting in I,q fitting, and not in log(I), log(q) fitting. Thank you, Irena Levin

1 0

Re: [Sasview-users] Extraction of form and structure factor
by Andrew Jackson 24 Nov '16

24 Nov '16

Dear Marco Richard is pointing to a new feature that should be in release 4.1 – it was added at the last Code Camp (in October), which was after the version 4 release. The separated form and structure factors will then appear in the Data Explorer under the data set you have fitted, similarly to how the fitted curve and residuals appear now (see screen shot). We are aiming to have the 4.1 release out in early 2017. Best Andrew ______________________________________ Andrew Jackson Instrument Scientist - Small Angle Scattering European Spallation Source, P.O Box 176, SE-221 00 Lund, Sweden Adjunct Associate Professor (Adjungerad Lektor) Physical Chemistry, Lund University, P.O. Box 124, SE-221 00, Lund, Sweden Phone: +46 46 888 3015 Mobile: +46 72 179 2015 E-mail: a<mailto:andrew.jackson@...631...>ndrew.jackson@...631... www.esss.se From: Richard Heenan <richard.heenan@...8...<mailto:richard.heenan@...8...>> Reply-To: Mailing list for users of SasView <sasview-users@...632...e.net<mailto:sasview-users@lists.sourceforge.net>> Date: Thursday 24 November 2016 at 13:44 To: "m.adamo15@...625...<mailto:m.adamo15@...625...>" <m.adamo15@...625...<mailto:m.adamo15@...625...>>, "users@...2...<mailto:users@...2...>" <users@...2...<mailto:users@...2...>> Subject: Re: [Sasview-users] Extraction of form and structure factor Dear Marco, Yes alas version 3 of sasview does not make it easy to save the S(Q). I suspect here that you have not allowed for the fitted flat background, which you will need to subtract, or set to zero and recalculate, before doing the division. In the new version 4 of sasview, the separate P(Q) and S(Q) are supposed to appear in the data structure on the left of the screen along with the total fit. I only say “supposed” in that in some of our very recent builds some other changes made this new feature disappear, and without going to look I am not sure of the current status on the public downloads. Hope this helps Richard Heenan From: Adamo, Marco [mailto:m.adamo15@...625...] Sent: 24 November 2016 11:13 To: users@...2...<mailto:users@...2...> Subject: [Sasview-users] Extraction of form and structure factor Dear SasView team, I have some corrected SANS data that I am fitting with SasView. The system is SDS 10% in 90% D2O/20% H2O v/v (file 093220_230). I was able to fit it with "EllipsoidModel”, “HayterMSAStructure” (file M2[093220_230_out.txt). Could you please explain me how to extract the structure and form factor? What I tried was to: 1. Fit the data with the above mentioned model 2. Remove the structure factor (just by deselecting it in SasView) to obtain ONLY the form factor (no more fit was performed, only the calculation) 3. Save the curve that is supposed to be a form factor 4. Divide the original data by the form factor to get the structure factor If I do the other way round, so I fit the data, remove the form, fix the structure factor (RED) and from it I get the form (PURPLE CIRCLES) by dividing the original data (BLUE) by the structure, I get this curve: [cid:image001.png@...629...] The problem is that I do not see any realistic form factor, as the structure is not completely removed at low q. Do you have any suggestions, please? Thanks, Marco

2 1

[Sasview-users] Extraction of form and structure factor
by Adamo, Marco 24 Nov '16

24 Nov '16

Dear SasView team, I have some corrected SANS data that I am fitting with SasView. The system is SDS 10% in 90% D2O/20% H2O v/v (file 093220_230). I was able to fit it with "EllipsoidModel”, “HayterMSAStructure” (file M2[093220_230_out.txt). Could you please explain me how to extract the structure and form factor? What I tried was to: 1. Fit the data with the above mentioned model 2. Remove the structure factor (just by deselecting it in SasView) to obtain ONLY the form factor (no more fit was performed, only the calculation) 3. Save the curve that is supposed to be a form factor 4. Divide the original data by the form factor to get the structure factor If I do the other way round, so I fit the data, remove the form, fix the structure factor (RED) and from it I get the form (PURPLE CIRCLES) by dividing the original data (BLUE) by the structure, I get this curve: [cid:image001.png@...624...] The problem is that I do not see any realistic form factor, as the structure is not completely removed at low q. Do you have any suggestions, please? Thanks, Marco

2 1

[Sasview-users] SasView 4.0 release
by Paul Butler 07 Oct '16

07 Oct '16

Dear SasView user, This is to let you know that 4.0 has now been released and is ready for use. Thanks to all of you who tried out the beta and gave us feedback which allowed us to catch several annoying issues. The development team is currently at code camp V at SNS in Oak Ridge working towards a release 4.1 as well as the GUI refactoring for a release 5.0 which will lead towards a future that includes the ability to run on cluster or cloud backends. We also have provided a brand new marketplace website, accessible at marketplace.sasview.org, where you can upload models to share with the community and/or download models from others that you can download to your plugins directory to run as is or edit and run. Currently only a couple of models from the release are available there but we hope to populate with all the models in the distribution of SasView. Those are also accessible from the SasView installation directory Thanks and we hope you enjoy the new version. The SasView Development Team

1 0

[Sasview-users] Please help test SasView 4.0 beta
by Paul Butler 08 Sep '16

08 Sep '16

Dear SasView user, You are receiving this email as a subscriber to the SasView user list and we wanted to let you know that version 4.0 is finally close to being released! While 4.0 has a number of bug fixes and feature requests added, the primary change is a complete rewrite of the way the fitting models are incorporated. The most important change simplifies the interface to SasView so that all built-in models can be used directly as custom plugins. This means, for example, that existing models can be altered for a specific purpose with relative ease. It also means that models that require C for speed are now supported --- in fact without the need for you to install a separate compiler! Just write your properly written and named C code in the plugins directory alongside the python module that calls it. Version 4.0 also makes all the infrastructure for built-in models, such as polydispersity functions, available to custom models. As part of this major refactoring, we also have taken advantage of the fact that all modern computers including laptops have a certain amount of GPU (graphical processor unit) power. As long as you have the OpenCL driver installed SasView will default to utilizing your GPUs when needed. OpenCl comes installed by default with Macs and many - but not all - PCs. Preliminary tests show that model-fitting, say, cylinders can get a 10-fold speedup. Should GPUs not be available on your machine, SasView will still utilize multiple CPUs. Finally, the model documentation has been extensively reviewed, checked, and edited. With such a massive restructuring there were bound to be “unexpected consequences” that came to light… and they did! The development team now believes that except for the tickets currently listed under Milestone 4.0.0 at http://trac.sasview.org/report/3 the beta 1 pre-release is robust enough for use in real applications. We are planning to have the full release available by the end of September. In the meantime, we would very much appreciate any feedback from your experience, and particularly bugs you find we may have missed. Thanks and we hope you enjoy the new version. The SasView Development Team

1 0

[Sasview-users] Regarding extraction or separate plot of structure factor, form factor and size distribution
by Vikram Raghuwanshi 10 Jun '16

10 Jun '16

Dear Sir/Madam, I am a new user to SASview. To fit a curve I used a sphere model with the structure factor and polydispersity. After fitting I would like to plot separately the structure factor and size distribution (which I used to fit the data). How I can plot and save data separately for the form factor, structure factor and size distribution in SASview ? Thank you Sincerely -- Dr. Vikram Singh Raghuwanshi Research Fellow Department of Chemical Engineering Faculty of Engineering 15 Alliance Lane, Clayton Campus, Monash University, VIC 3800 Australia Tel: +61 399055088 Email: vikram.raghuwanshi@...622...

2 1