Use monte carlo integration for high dimensional polydispersity integrals

Instead of increasing the number of PD loops in the kernel code, switch to Monte Carlo integration with importance sampling when more than 4 dimensions. Do not extend the code to contain one more loop. Not only will this be slower, but it will increase code size and possibly reduce the number of q values that can be evaluated in parallel.

Many 1D models are computed with numerical integrals over all orientations; these integrals can be incorporated into the Monte Carlo. Could also incorporate sampling from the resolution function.

Accuracy can be controlled by increasing the number of samples.

Migrated from http://trac.sasview.org/ticket/796

{
    "status": "new",
    "changetime": "2016-10-25T18:21:55",
    "_ts": "2016-10-25 18:21:55.387374+00:00",
    "description": "Max 4 polydisperse parameters. Need warnings in GUI if more are selected.\n\nInstead of increasing the number of PD loops in the kernel code, switch to Monte Carlo integration with importance sampling when more than 4 dimensions. Do not extend the code to contain one more loop. Not only will this be slower, but it will increase code size and possibly reduce the number of q values that can be evaluated in parallel.\n\nMany 1D models are computed with numerical integrals over all orientations; these integrals can be incorporated into the Monte Carlo.  Could also incorporate sampling from the resolution function.\n\nAccuracy can be controlled by increasing the number of samples.",
    "reporter": "pkienzle",
    "cc": "",
    "resolution": "",
    "workpackage": "SasView Bug Fixing",
    "time": "2016-10-25T18:21:38",
    "component": "sasmodels",
    "summary": "Polydispersity limited to 4 dimensions",
    "priority": "major",
    "keywords": "",
    "milestone": "SasView Next Release +1",
    "owner": "",
    "type": "enhancement"
}

SasView / sasmodels

Use monte carlo integration for high dimensional polydispersity integrals #132