Follow-up Comment #6, bug #22024 (project relax):
This map shows the total failure op minimising the chi2 space, with reasonable
experimental parameters.
It is very very weird, that the optimized point lies so far down the edge.
This happens when dw=1 or less for this setup.
Why is this???
The grid value of pA was close the right answer.
But the minimisation went down to 0.5 !
GRID=0.975 MIN=0.500 CLUST=0.500 SET=0.993
sfrq_1 = 599.8908617*1E6
time_T2_1 = 0.06
ncycs_1 = [28, 4, 32, 60, 2, 10, 16, 8, 20, 50, 18, 40, 6, 12, 24]
sfrq_2 = 499.8908617*1E6
time_T2_2 = 0.04
ncycs_2 = [20, 16, 10, 36, 2, 12, 4, 22, 18, 40, 14, 26, 8, 32, 24, 6, 28 ]
r20_key_1: 25.0, r20_key_2: 24.0
Space created with:
'kex': 2200.0, 'pA': 0.993, 'dw': 1.0
Params for dx map is
['dw', 'pA', 'kex']
Point creating dx map is
[1.0, 0.99299999999999999, 2200.0]
Minimises point in dx map is
[0.165081668808779, 0.50002449305070162, 2255.0279528019505]
########################
Generated data with MODEL:NS CPMG 2-site expanded
Analysing with MODEL:CR72.
########################
For spin: ':1@N'
CR72 Ala :1@N r2 599.9 GRID=25.041 MIN=25.000 CLUST=25.000 SET=25.000
RELC=0.000
CR72 Ala :1@N r2 499.9 GRID=24.029 MIN=24.000 CLUST=24.000 SET=24.000
RELC=0.000
CR72 Ala :1@N pA GRID=0.975 MIN=0.500 CLUST=0.500 SET=0.993 RELC=0.986
###################################
WARNING: pA Have relative change above 0.05, and is 0.9859.
###################################
CR72 Ala :1@N dw GRID=0.500 MIN=0.165 CLUST=0.165 SET=1.000 RELC=5.058
###################################
WARNING: dw Have relative change above 0.05, and is 5.0576.
###################################
CR72 Ala :1@N kex GRID=2000.800 MIN=2255.028 CLUST=2255.028 SET=2200.000
RELC=0.024
(file #20698, file #20699, file #20700)
_______________________________________________________
Additional Item Attachment:
File name: CR72_10_dw_error_map_1_N.png Size:176 KB
File name: CR72_10_dw_error.py Size:24 KB
File name: CR72_10_dw_error_disp_1_N.png Size:7 KB
_______________________________________________________
Reply to this item at:
<http://gna.org/bugs/?22024>
_______________________________________________
Message sent via/by Gna!
http://gna.org/
[bug #22024] Minimisation space for CR72 is catastrophic. The chi2 surface over dw and pA is bounded.