Changes between Version 3 and Version 4 of CodeCampIX/BreakOut1


Ignore:
Timestamp:
Mar 31, 2019 9:48:44 AM (5 years ago)
Author:
butler
Comment:

Legend:

Unmodified
Added
Removed
Modified
  • CodeCampIX/BreakOut1

    v3 v4  
    1 == Sasmodels Enhancement Discussion == 
     1== Sasmodels Enhancement Discussion Agenda == 
    22* From Roadmap 
    33  * reparametrization of models -  
     
    3838 * How do we do this? 
    3939  * Can we make a good interface for the user to do parameter mapping  
    40   * Wojciech - the optimiser might have problems with that. 
    41   * Non-fittable parameters 
    42    * Involved with a constraint 
    43    * Calculated outputs 
    44    * See SLD profiles, see 3D shape 
    45   * Narayan: Make sure that it is clear that constrained parameters are visible and clear that they are constrained. 
     40   * Wojciech - the optimiser might have problems with that if we are not careful. 
     41  * RKH: Would need non-fittable (static) parameters 
     42   * PDB: this is also been needed with constraints 
     43   * AJJ: Also been needed in calculated outputs 
     44   * "There is a ticket for that" 
     45   * Also other static output plots (see SLD profiles, see 3D shape) would be nice .. though off-topic here. 
     46  * Narayan: Make sure that all constrained parameters (fitable or not) are visible and that it is clear that they are constrained. 
    4647  * PDB: Actually should be easy to provide an on the fly reparameterization: Use a GUI tool like the Add Model Editor to provide parameters, any constraints and their relationship to the existing parameters.  Then "create the code" by editing the model file: adding the conversion math at the beginning of the function. Essentially the infrastructure and a first order proof of concept already. 
    47   * Start with a simple test, then maybe talk to Paul Kienzle 
    48   * Think about GUI 
     48AGREED - Start with a simple test, then maybe talk to Paul Kienzle 
    4949 
    5050=== Include more complex structure factors === 
     
    5252  * We don’t have complex amplitudes 
    5353   * Use case : Janus particles - but can cheat it with end capped cylinder 
    54  * Beta 
    55  * Locally monodisperse 
    56  * More ways of dealing with structure factors 
    57   *pyPRISM - will be implemented in SasView by Tyler Martin 
    58  * Discussion of SasFit / SasView 
    59  * Models 
    60  * OZ Solver 
    61  * More ways of dealing with polydispersity 
    62  * Orientation dependent Structure factors that depend on form factor (amplitude) 
    63 Write analytical ones 
    64 Yun - only good for hard sphere and spherical particles  
    65 Yun - Continuous distributions in OZ framework is hard 
    66 Yun - should go for numerical solutions for more complex interaction potentials 
    67 Yun - for e.g. mixture of latex and gold, need cross terms, so need complex amplitudes 
     54 * Discussion of !SasFit / !SasView 
     55  * More ways of dealing with structure factors 
     56   * Orientationally dependent structure factors 
     57    * e.g. Beta - Now in !SasView 
     58   * Structure factors that depend on form factor (amplitude) 
     59   * More ways of dealing with polydispersity 
     60    * Beta - Now in !SasView 
     61    * Locally monodisperse 
     62  * Other Models 
     63  * OZ Solver 
     64 * pyPRISM - will be implemented in !SasView by Tyler Martin 
     65 * Probably for most cases nee to just write analytical ones 
     66  * Yun - Most cases will only be good for hard sphere and spherical particles  
     67  * Yun - Continuous distributions in OZ framework is hard 
     68  * Yun - should go for numerical solutions for more complex interaction potentials 
     69  * Yun - for e.g. mixture of latex and gold, need cross terms, so need complex amplitudes 
    6870 * Sylvain - Finish decoupling approximation first? For spherical. 
    69 ALL - AGREED last point 
     71AGREED - Finish decoupling and Prism before talking more.  This then becomes lower priority, since what is left is both quite complicated and not easy to be general (i.e. each solution will work for a very small number of cases).  
    7072 
    7173=== Include material from Scatter === 
    72 Andrew has had student working on it. Difficult. Needs hypergeometric functions. 
    73 MLZ have been working on it too - a few simple models. 
    74 They have some proof of concept in Born Again 
    75 Need to work out API to work with SasView 
    76 Plan for this ... 
    77 Aim for next Code Camp to have BornAgain team integrate their code. 
    78 Check SasFit - Joachim may have implemented already. 
     74 * Andrew has had student working on this. It is quite difficult. Key is the use of hypergeometric functions.  This might also be good for integration speeds in general (see below). 
     75 * MLZ have been working on it too for a few simple models. 
     76  * They have some proof of concept in Born Again 
     77  * Need to work out API to work with SasView 
     78  * Plan for this ... 
     79   *Aim for next Code Camp to have BornAgain team integrate their code. 
     80  *Check !SasFit - Joachim may have implemented already. 
    7981 
    8082=== 1D oriented data ... === 
    81 How to fit a slice from a 2D dataset 
    82 Do we do it? How? 
    83 Make special 1D models? 
    84 Or cut 2D models? 
    85 Speed … ? Need to look at the speed ...= 
    86 User interface … ? 
    87 How to show the slicer options to the user? 
    88 Narayan / Sylvain to find some example data. 
    89 Maybe start with simulation and treat as real data to test process 
    90 AJJ to make some simulated data and share with everyone. 
     83 * How to fit a slice from a 2D data sets (oriented objects, magnetic scattering, etc.)  
     84 * Should we do it? 
     85  * If we are going to provide live access to fast computers 
     86  * If not too hard maybe do anyway .. a number of users with good cases 
     87 * How would we do it? 
     88  * Make special 1D models? 
     89  * Or cut 2D models? 
     90 * Speed … ? 
     91  * Need to look at the speed ...= 
     92 * User interface … ? 
     93  * How to show the slicer options to the user? 
     94 * Narayan / Sylvain to find some example data. 
     95  * Maybe start with simulation and treat as real data to test process 
     96 * AJJ to make some simulated data and share with everyone. 
    9197 
    9298=== Project to do benchmarking of optimisers === 
    93 * Wojciech : Idea is that there is a summer student at ISIS working with Anders 
    94 Mantid have been testing their optimisers seeing how they perform on different classes of problems.  Want to do something similar for SasView 
    95  * Test SasView optimisers against their defined problems 
    96  * Identify classes of problems that would be representative for the SANS community 
    97 * AJJ - need to know the answer? 
    98 * AJJ - how is SANS mathematically different from other test cases? 
    99 * RKH - What about things where we have models that have multiple possible simulations? 
     99 * Wojciech : Idea is that there is a summer student at ISIS working with Anders 
     100  * Mantid have been testing their optimisers seeing how they perform on different classes of problems.   
     101  *Want to do something similar for !SasView 
     102   * Test !SasView optimisers against their defined problems 
     103   * Identify classes of problems that would be representative for the SANS community 
     104  * AJJ - need to know the answer? 
     105  * AJJ - how is SANS mathematically different from other test cases? 
     106  * RKH - What about things where we have models that have multiple possible simulations? 
    100107 
    101108=== Computation speed and integration options and approaches ===