| T O P I C R E V I E W |
| BennettS |
Posted - 05/02/2018 : 07:55:36 AM
Origin Ver. and Service Release (Select Help-->About Origin): Origin 2018G (64-bit)
Operating System: Windows 7
Hello,
I'm using the following LabTalk script to facilitate my analysis of ultrasonic data, which runs mostly as intended:
page.active = 1;
//zeropadding for FFT
csetvalue col:=col(A)[1001:10000] formula:="col(a)[1000]+((i-1000)*(col(a)[1000]-col(a)[999]))";
csetvalue col:=col(B)[1001:10000] formula:="0";
//FFT
fft1 col(B) win:=hanning re:=0 im:=0 ap:=0 ri:=0 db:=0 psd:=0;
//preparation for my question, essentially an user input to create an identifier
string comment$="Kommentar";
getn (Comment) comment$;
range cc = 1!col(B);
cc.Comment$ = comment$;
//plot FFT and fit
page.active$ = "FFTResultData1";
plotxy (1,2) plot:=200;
nlbegin 1 func:=Lorentz_3 nltree:=ParamTree;
nlfit;
nlend 1;
1)a side question here: when doing the fit "by hand" using the dialogue box, the graph will show the green lock in the upper left corner, making it possible to rerun/modify the fitting if necessary. With the script, the lock is missing. Is it possible to change that?
2)the fit function I’m using is a custom one with 9 parameters (A1-A3, w1-w3, f1-f3). I have multiple worksheets with my results from different measurements and will be running the script on each one one by one. Now I would like to add something to the script to write the fit parameters in Y columns and the identifier (the range cc) as X into a prepared workbook eg. named “FitResult”, so each run of the script over a new data workbook would add the identifier and parameters in the next free line of “FitResults”
I think I can access the identifier with the range I declared and the parameters probably from nltree somehow, but am at a loss as to how then get them linewise into my result workbook.
Thanks for your help. |
| 4 L A T E S T R E P L I E S (Newest First) |
| BennettS |
Posted - 05/08/2018 : 06:17:17 AM
Thank you for the hint, I got the script to do everything I wanted now.
The cell notation seems to be more forgiving in regards to a variable input than col()[] and csetvalue can apparently be used to put a string into a cell when directly assigning it does not work. Well, things learned...
For reference, the working code:
//Part 1
newbook name:="Fit" sheet:=1 option:=lsname chkname:=1;
range rWfit = [Fit]Sheet1!;
rWfit.nCols = rWfit.nCols + 14;
col(A)[L]$ = Kommentar;
col(B)[L]$ = A1;
col(C)[L]$ = A2;
col(D)[L]$ = A3;
col(E)[L]$ = w1;
col(F)[L]$ = w2;
col(G)[L]$ = w3;
col(H)[L]$ = f1;
col(I)[L]$ = f2;
col(J)[L]$ = f3;
col(K)[L]$ = A1rel;
col(L)[L]$ = A2rel;
col(M)[L]$ = A3rel;
col(N)[L]$ = att1;
col(O)[L]$ = att2;
col(P)[L]$ = att3;
csetvalue col:=col(K) formula:="Col("A1")/(Col("A1")+Col("A2")+Col("A3"))" recalculate:=1;
csetvalue col:=col(L) formula:="Col("A2")/(Col("A1")+Col("A2")+Col("A3"))" recalculate:=1;
csetvalue col:=col(M) formula:="Col("A3")/(Col("A1")+Col("A2")+Col("A3"))" recalculate:=1;
csetvalue col:=col(N) formula:="Col("w1")/Col("A1")" recalculate:=1;
csetvalue col:=col(O) formula:="Col("w2")/Col("A2")" recalculate:=1;
csetvalue col:=col(P) formula:="Col("w3")/Col("A3")" recalculate:=1;
//part 2
page.active = 1;
csetvalue col:=col(A)[1001:10000] formula:="col(a)[1000]+((i-1000)*(col(a)[1000]-col(a)[999]))";
csetvalue col:=col(B)[1001:10000] formula:="0";
fft1 col(B) win:=hanning re:=0 im:=0 ap:=0 ri:=0 db:=0 psd:=0;
string comment$="Kommentar";
int ii = 1;
getn (Comment) comment$
(Index) ii;
range cc = 1!col(B);
cc.Comment$ = comment$;
page.active$ = "FFTResultData1";
plotxy (1,2) plot:=200;
nlbegin 1 func:=Lorentz_3 nltree:=ParamTree;
nlfit;
nlend 1 2;
csetvalue col:=[Fit]Sheet1!col(A)[ii] formula:="comment$";
[Fit]Sheet1!Cell($(ii),2) = ParamTree.A1;
[Fit]Sheet1!Cell($(ii),3) = ParamTree.A2;
[Fit]Sheet1!Cell($(ii),4) = ParamTree.A3;
[Fit]Sheet1!Cell($(ii),5) = ParamTree.w1;
[Fit]Sheet1!Cell($(ii),6) = ParamTree.w2;
[Fit]Sheet1!Cell($(ii),7) = ParamTree.w3;
[Fit]Sheet1!Cell($(ii),8) = ParamTree.f1;
[Fit]Sheet1!Cell($(ii),9) = ParamTree.f2;
[Fit]Sheet1!Cell($(ii),10) = ParamTree.f3;
|
| yuki_wu |
Posted - 05/06/2018 : 10:22:43 PM
Hi,
I think this example could help:
for(int ii = 1; ii < 10; ii++)
{
[Book1]Sheet2!Cell($(ii),2) = 10;
}
Regards,
Yuki
OriginLab
|
| BennettS |
Posted - 05/03/2018 : 04:13:30 AM
Thanks for your help so far, #1 is solved, seems I missed the correct option notation for nlend.
For #2, I’ve thought of a (clunky) workaround following your hint. As I’m prompting for an identifier anyway, I could ask for a row index in the same step and then use a variable in my cell assignment, eg.
string comment$="Kommentar";
r = 1;
getn (Comment) comment$
(Index) r;
range cc = Sheet1!col(B);
cc.Comment$ = comment$;
//fit part from post above runs
[“FitResults”]Sheet1!col(1)[r] = comment$;
[“FitResults”]Sheet1!col(2)[r] = ParamTree.A1;
[“FitResults”]Sheet1!col(3)[r] = ParamTree.A2;
[“FitResults”]Sheet1!col(4)[r] = ParamTree.A3;
//and so on for the remaining parameters
the problems that I’ve now run into are:
1) using r as row index apparently is not the correct way to create a row variable
2) comment$ can’t be assigned as a cell value, I’m guessing a data type conflict
3) I’m missing something basic about cell designation,
//this works fine, with the right sheet active of course
col(2)[1] = ParamTree.A1;
//these don’t
Sheet1!col(2)[1] = ParamTree.A1;
[“FitResults”]Sheet1!col(2)[1] = ParamTree.A1;
|
| yuki_wu |
Posted - 05/02/2018 : 11:26:45 PM
Hi,
#1 Add a parameter at the end of the nlend X-Function, for example:
nlend 1 2; //Recalculate Mode: Manual
nlend 1 1; //Recalculate Mode: Auto
More info please take a look at this page:
https://www.originlab.com/doc/X-Function/ref/nlend
#2 I suppose you hope to put the value of parameters to the specified cells in a worksheet, right? If so, you can use a range to access a single cell or block cells, for example:
// cell(2,1), row2 of col(1)
range aa = 1[2];
// cell(1,1) to cell(10,3)
range bb = 1[1]:3[10];
More info about Range:
https://www.originlab.com/doc/LabTalk/guide/Range-Notation#Block_of_CellsRange.2C_Block_of_Cells
Hope it helps.
Regards,
Yuki
OriginLab
|
|
|
