| Author |
 Topic  |
|
|
gleeblejim
USA
33 Posts |
 Posted - 09/18/2012 : 09:07:45 AM
|
Origin Ver. 8.6 and Service Release 3 (Select Help-->About Origin):
Operating System: Win 7
Good Day Origin Crew!
I'm looking to find the time delay between two sets of data, in this case Distance (measured) and PDistance(ideal). The tests will be controlled so the only thing that will change from test to test is Distance. Is there any way in Set Column Values to tell Origin to average the time difference between two columns, for instance when Distance = 2,4,6,8 then find Average(Time of PDistance at # - Time of Distance at #)?
Thank you,
Jim |
|
|
Drbobshepherd
USA
Posts |
Posted - 09/18/2012 : 3:56:03 PM
|
Jim,
I am not sure, but it looks like from your example that you are describing a time measurement for a fixed set of distances. Are "Distance" the independent variable, and "Predicted Time at Distance" and "Measured Time at Distance" your dependent variables?
I'm just guessing, but I think you are asking how to use Set Column Values to average a number of datasets and subtract the average from a dataset of predicted values. If you know the number of datasets to average, it isn't difficult. For example, say the following table is your data:
d tPred tMeas1 tMeas2 tMeas3
2 5.000 4.2040 4.5160 4.6230
4 7.777 7.6050 7.5430 7.6120
6 9.500 8.9700 9.0010 9.5960
8 11.00 10.502 10.999 10.755
Add a column. Set Column Values so that Col(F)=
Col(B)-((Col(C)+Col(D)+Col(E)/3))
Apply an appropriate label to this column and you are done. The results should look something like this:
d tPred tMeas1 tMeas2 tMeas3 Avg(tP-tM)
2 5.000 4.2040 4.5160 4.6230 0.55233
4 7.777 7.6050 7.5430 7.6120 0.19033
6 9.500 8.9700 9.0010 9.5960 0.31100
8 11.00 10.502 10.999 10.755 0.24800
If the number of measured datasets is unknown, you should write a LabTalk script to get the number and calculate the average in the advanced Before-Formula-Script window in the Set-Column-Values dialog.
I hope this answered your question?
DrBobShepherd |
 |
|
|
gleeblejim
USA
33 Posts |
Posted - 09/18/2012 : 4:09:03 PM
|
Sorry for not clarifying in the original post. In my software Time is always acquired as the X value and everything else is Y. The data file would look like:
Time(X) PDistance(Y) Distance(Y)
(sec) (mm) (mm)
0 0 0
1 1 0.5
2 2 1
3 3 2
4 4 3
5 4 4
And I would like to make a template that has one of the columns automatically find the time delay for the two columns. The attached image shows the area I'm interested in. |
|
|
|
Drbobshepherd
USA
Posts |
Posted - 09/19/2012 : 12:20:08 PM
|
OK. Try this.
1. Add a column to the 3 you already have. Longname it "Delay".
2. Select the column, and open the Set Column Values dialog.
3. Enter the following in the Before-Formula-Script window:
interp1 ix:=Col(C) iy:=(Col(B), Col(A)) ox:=Col(D); // Interpolate t value on PD curve for each D value.
Col(D)=Col(D)-Col(A); // Calculate delay.
avg=mean(Col(D)); // Calculate mean delay.
Col(D)[C]$=Mean= $(avg,.3); // Type value in Comments, format to 3 decimal places.
4. Click on OK.
I hope this works for you. If it does, you may wish to save it as an Analysis Workbook for future processing.
DrBobShepherd |
|
|
|
Hideo Fujii
USA
1582 Posts |
Posted - 09/19/2012 : 2:32:05 PM
|
Hi Jim,
Maybe I'm lazy, but if I were, I will fit your two plots with a sigmoidal function, e.g., Boltzmann (independently, or globally with shared parameters); Then, you can subtract the x coordinates of the centers.
It would work even when the data is noisy.
--Hideo Fujii
OriginLab |
Edited by - Hideo Fujii on 09/19/2012 2:33:50 PM |
|
|
| |
Topic |
|
|
|
Time Delay