Ticket #713 (closed defect: fixed)

Opened 7 years ago

Last modified 7 years ago

Stretched exponential fit results don't agree with TRI2

Reported by: aivar Owned by: aivar
Priority: major Milestone: slimplugin1.2
Component: slim-plugin Severity: serious
Keywords: Cc:
Blocked By: Blocking:

Description

Using same RLD+LMA fit as #697, (146, 230) of 100611-YG-256.sdt, Gaussian noise:

Single exponential results agree, but stretched exponential differ:

TRI2: A 93.23 T 1.78 H 1.15 Z 0.77 X2 0.858475

SLIM Plugin: A 87.63362 T 1.911811 H 100.0 Z 1.1207124 X2 22.582573

Change History

comment:1 Changed 7 years ago by aivar

  • Status changed from new to assigned

comment:2 Changed 7 years ago by aivar

Once this is working better there will still be an issue with the tau:

See TRfitting.c DoEcfFitting()

Report the average tau for stretched if (fitType[fit_type].fitType == FIT_STRETCHED) {

convert tau into <tau> param[2] = param[2]*param[3] * Gamma(param[3], NULL);

}

comment:3 Changed 7 years ago by aivar

  • Blocking 674 removed

Removed blocking #674 so I could close that ticket. This is a very specific instance of TRI2/SLIM Plugin disagreement unlike the others.

I will address this bug when I work on SLIM Curve; I'm planning to add a Gamma function there.

comment:4 Changed 7 years ago by aivar

  • Milestone changed from slimplugin1.1 to slimplugin1.2

comment:5 Changed 7 years ago by aivar

  • Milestone changed from slimplugin1.2 to slimplugin1.1A

comment:6 Changed 7 years ago by aivar

  • Status changed from assigned to closed
  • Resolution set to fixed

Now that I have GCI_gamma this is easy to fix.

comment:7 Changed 7 years ago by aivar

Verified that for 100611-YG-256.sdt, pixel (144, 230), cursors 1.3184/1.8066/11.2305, stretched exponential, maximum likelihood, RLD+LMA I get:

TRI2 Z 1.225 A 58.804 tau 1.495E+0 h 1.415

SP Z 1.225 A 58.804 tau 1.495 h 1.415

Note: See TracTickets for help on using tickets.