| Author |
 Topic  |
|
|
Belsinga
Netherlands
28 Posts |
 Posted - 10/25/2010 : 08:17:32 AM
|
Origin Ver. and Service Release (Select Help-->About Origin): 8 SR 6
Operating System: Win XP
Dear all,
I need to know how I can fit a dataset with nlsf, where the error in x is situated in a column. When I dedicate this column as "X-error", the script delivers the following error:
"External call error"
If this column (with x errors) is used as regular Y values, I can fit the data, but it uses the column as "Y-error". Can someone help me?
Thanks in advance
(in the code, FT_Q2 is the name of the column where my values for x-error are.)
******************************************
code:
nlsf.cleanupfitdata();
nlsf.func$ = Allometric1;
a=1;
b=1;
nlsf.numfitsets = 1;
nlsf.numIndep = 1;
nlsf.dataBegin = 1;
nlsf.dataEnd = vsA;
limit FT_Cell3;
nlsf.fitdata1$ = FT_Cell3;
nlsf.funcx$ = FT_Transmit;
nlsf.w$ = FT_Q2;
nlsf.wtype = 1;
nlsf.xPoints = vsA;
nlsf.fit(100);
nlsf.makecurve(func);
nlsf.pasteParams(R);
nlsf.paramWks(Param_cell3); |
|
|
larry_lan
China
Posts |
Posted - 10/26/2010 : 01:53:22 AM
|
Hi:
Origin neither supports fitting X-Error as weight, nor fitting dataset that set as X-Error. If you want to keep FT_Q2 as "X-Error" column type, maybe you can duplicate it as "Y" (or Y-Error) column type and use data in this new column as weights.
Thanks
Larry
OriginLab |
 |
|
|
Belsinga
Netherlands
28 Posts |
Posted - 10/28/2010 : 09:01:45 AM
|
Ok,
But what about leaving the x-error column in the worksheet (still defined as "x-err"). Won't Origin pick this up automatically as an uncertainty?
Thanks,
Boudewijn |
|
|
| |
Topic |
|
|
|
X-error in NLSF