| T O P I C R E V I E W |
| dean.pask |
Posted - 11/01/2005 : 6:29:36 PM
Origin Version (Select Help-->About Origin): 7.5
Operating System: xp
Hi all,
My program with large amounts of help from Mike is below which works amazing.. I have found that my FFTs are skewed and require interpolating then curve fitting (non linear curve fitting - peak functions - gauss)... my question as ever is how do I do this in my code?
Super thanks in advance again.
Dean
%L=%H; // save its name
wks.col1.type=4;
worksheet -s 2 0 2 0;
worksheet -p 200 Origin;
layer -a;
window -t W FFT;
%F=%H; // save its name
page.label$=FFT of %L; // identify Labview data in label
page.title=3; // show label
fft.reset();
fft.forward = 1;
fft.forward.timeData$ = %L_A;
fft.forward.tdelta = %L_A[2] - %L_A[1];
fft.forward.realData$ = %L_B;
fft.output.samplingdata$ = %F_Freq;
fft.output.realdata$ = %F_Real;
fft.output.imagdata$ = %F_Imag;
fft.output.ampdata$ = %F_r;
fft.output.phasedata$ = %F_Phi;
fft.output.powerdata$ = %F_Power;
fft.real = 1;
fft.normalize = 1;
fft.shifted = 1;
fft.windowing = 3;
fft.spectrum = 1;
fft.unwrap = 1;
fft.forward();
worksheet -s 4 0 4 0;
worksheet -p 200 Origin;
%P=%H; // save graph name
layer1.y.from = 0;
layer1.x.from = 2E8;
layer1.x.to = 3E9;
layer1.y.to = 0.5;
nPeaks=5;
%B=%F_r;
win -t D; // create temporary wks
%W=%H; // save its name
wks.addcol(Pki); // add col for peak indices
curve.pickpeaks.simple=1;
curve.reset();
curve.data$=%B;
curve.peakIndex$ = %W_Pki;
curve.pickPeaks.rectHeight = 0.005;
curve.pickPeaks.rectWidth = 0.005;
curve.pickPeaks.minHeight = 0.005;
curve.pickPeaks(1);
get %W_Pki -e end;
if(end==0) {
type -a No peaks were found.;
win -cd %W;
return;
};
nn=0;
for(ii=1;ii<=end;ii++)
{
index = %W_Pki[ii];
xx = xvalue(index,%B);
if( xx<0 ) continue;
nn++;
%W_A[nn] = xx;
%W_B[nn] = %B[index];
};
type -a $(nn) peak/s were found;
if(nn>nPeaks)
{
sort -wd %W %W_B;
set %W -er nPeaks;
sort -w %W %W_A;
};
if(nn>0)
{
win -a %F; // activate FFT wks
wks.addCol(Pki);
wks.addCol(Pkx);
wks.addCol(Pky);
copy %W_Pki %H_Pki;
copy %W_A %H_Pkx;
copy %W_B %H_Pky;
};
win -cd %W; // delete temporary wks
win -a %P; // activate FFT graph
%W=%[%C,'_']; // get name of the FFT wks from active dataplot
if( exist(%W_Pkx,1)==0 ) {
ty -a No peaks!;
return;
};
get %W_Pkx -e nPeaks; // # of peaks
if( nPeaks==0 ) {
ty -a No peaks!;
return;
};
loop (ii,1,nPeaks) {
xx=%W_Pkx[ii]; // pk position
yy=%W_Pky[ii]; // pk height
label -s $(xx); // create text label
doc -e G {
// find its name - will be the first match encountered
if("%[%B,5]"=="text") break;
};
%B.x = xx; // move center of label to X = xx
%B.y = yy + %B.dy; // move center to Y = yy + labelHeight
};
|
| 1 L A T E S T R E P L I E S (Newest First) |
| Mike Buess |
Posted - 11/02/2005 : 09:11:51 AM
Hi Dean,
If linear interpolation is sufficient you can use the Table() function. Not sure where you want to do this but assume your FFT data are in cols Freq(X) and r(Y) of wks FFT1. Create a new wks (Data2), fill A(X) with the new frequency values and run this command...
Data2_B = Table(FFT1_Freq, FFT1_r, Data2_A);
If more sophisticated interpolation is required you can use the Curve Spline Interpolate function here...
http://www.originlab.com/index.aspx?s=8&lm=243
Mike Buess
Origin WebRing Member |
|
|
