Opened 8 years ago

Closed 8 years ago

#318 closed defect (fixed)

GA MC statistics calculation wrong

Reported by: titus_franzmann Owned by: gegorbet
Priority: normal Milestone: future
Component: ultrascan3 Version:
Keywords: review Cc:

Description

The Molecular weight and concentration show negative values for the confidence values. And, though I did yet not calculate the confidence levels by hand, the 9.8 for species 1 seems way too high. Same applies to species 2.

Summary:

Solute 1:
Molecular weight: 6.7541e+01 (-2.1323e+03, 9.8654e+03)
Concentration: 2.3492e-02 (-8.3238e-03, 8.8102e-02)
Total Concentration: 2.5928e-02
Sedimentation Coeff.: 1.0035e-13 (9.7895e-14, 1.1243e-13)
Diffusion Coeff.: 1.2759e-06 (1.2621e-06, 1.3408e-06)
Frictional Ratio, f/f0: 1.3274e+00 (1.3225e+00, 1.4168e+00)

Solute 2:
Molecular weight: 9.9349e+01 (-2.4783e+03, 1.5314e+04)
Concentration: 1.4643e-01 (-8.5524e+00, 1.8279e+01)
Total Concentration: 7.2946e-01
Sedimentation Coeff.: 1.5246e-13 (1.5171e-13, 1.5420e-13)
Diffusion Coeff.: 1.2342e-06 (1.2201e-06, 1.2722e-06)
Frictional Ratio, f/f0: 1.2336e+00 (1.2296e+00, 1.2572e+00)

Relative Concentrations:

Total concentration: 0.755387 OD
Relative percentage of Solute 1: 3.43241 %
Relative percentage of Solute 2: 96.5676 %

Attachments (1)

buggy ga mc result.svg (44.5 KB) - added by titus_franzmann 8 years ago.

Download all attachments as: .zip

Change History (4)

Changed 8 years ago by titus_franzmann

comment:1 Changed 8 years ago by demeler

  • Owner changed from bdubbs to gegorbet

I agree, this seems very strange, must be a calculation error in the stats routine. Gary: You could look at the US2 code for this, it does the stats correctly. You should be able to lift the logic directly from there for US3.

comment:2 Changed 8 years ago by gegorbet

  • Keywords review added

Version 1207 corrects some statistics calculations, by including use of weighted averaging in the way that US2 did.

comment:3 Changed 8 years ago by dzollars

  • Resolution set to fixed
  • Status changed from new to closed
Note: See TracTickets for help on using tickets.