| T O P I C R E V I E W |
| liana.shirmane |
Posted - 01/23/2014 : 04:20:08 AM
Hello!
I'm using Origin Pro 8.0 and have a question.
I have graph with Voltage on Y scale and on X - time.
Also I have calibration data table Voltage and Transmittance (%). Plot it and received Transmittance (Y) and Volttage (X). Than I have done interpolation and received table with Voltage and Transmittance in %.
Now I need to substitute Voltages on Transmittance and get as result graph with Transmittance on Y and time on X.
Couldn't find any tutorial.
Can someone please help me?
PhD student |
| 6 L A T E S T R E P L I E S (Newest First) |
| Drbobshepherd |
Posted - 01/27/2014 : 4:32:25 PM
Liana,
(1) In general, it can make a difference in the interpolated result if you reverse the designations of your dependent variable y and the independent variable x in your calibration function. Because y is a function of x doesn't mean x is necessarily a function of y.
Also, the Origin interpolation routines need to know in the function to be interpolated, which column is x and which column is y. If the designations are incorrect, you will get x-values instead of y-values in the result.
(2) In Step 4, I suggested setting "Col(E) = Col("Tc")(Col("Vm"))". This is a LabTalk command telling Origin to perform a piece-wise linear interpolation of the Tc function at the current Vm(i) value. The Set Column Value routine loops this command for i=1 to i=max cell number in Col("Vm").
This is an an example of the difference between using "()" and "[]" to set column values.
"Col(C)[1]=Col(B)[3]" means copy the value in Column(B), Row(3) to Column(C) Row(1).
"Col(C)[1]=Col(B)(3)" means set the value of Col(C), Row(1) to a linear interpolated value of the Col(B) function for x=3. Origin will attempt to identify the correct X-column associated with Col(B) to calculate the line. If there is only one X-column in the sheet, Origin should not have any difficulty.
Be careful you do not perform another Set Column Values after you change any column designations to X; this could confuse Origin and give you incorrect results. Set the Column Value manually, not automatic.
Linear interpolations are quick and easy, but spline-based interpolations produce smoother curves and usually smaller errors. If you want to try using splines, use the "Analysis: Mathematics: Interpolate/Extrapolate Y from X" dialog box as suggested by lkb0221, and select one of the splines instead of "Linear" for your Method.
This could be a good opportunity to research, experiment, and learn about interpolation.
DrBobShepherd |
| lkb0221 |
Posted - 01/27/2014 : 09:40:15 AM
Hi, Liana
Assume you have four columns: A(Time (x)), B(Voltage measured (y)), C(Voltage (x)), D(Transmittance (y))
Please highlight col(C) and col(D) and select "Analysis: Mathematics: Interpolate/Extrapolate Y from X", set "X Values to Interpolate" be col(B), then click OK. (Input is col(C) and col(D))
The created column is the interpolated col(D) based on col(B).
Zheng
OriginLab |
| liana.shirmane |
Posted - 01/26/2014 : 12:00:46 PM
Once again, thanks for help.
Unfortunately, I wasn't correct in my first post.
As calibration I have - Volts (Y) and Transmittance (X). We measured Voltage, but Absorption (and later Transmittance we knew).
lkb0221, so I have dataset
Columns A (Transmittance (x)), B (Voltage (y)), C (Time (x)), D (Voltage measured (y)).
If I select "Analysis: Mathematics: Interpolate/Extrapolate Y from X"
What I should choose as
X Values to Interpolate - ?
Input - ?
Drbobshepherd, thank for such detailed description.
But I have some more questions :
1)is there any difference in steps/columns, if my Tc is X and Vc is Y? As I think it isn't a big difference, as a result we get anyway calibration curve.
2)How this generation of linear-interpolated values works? As my calibration curve consist only of 8 points,but my measured data is 2000 points, that's why I tried to interpolate first, to get more points.
PhD student |
| liana.shirmane |
Posted - 01/24/2014 : 02:36:40 AM
Thanks everyone, will try!
PhD student |
| Drbobshepherd |
Posted - 01/23/2014 : 11:21:09 AM
Liana,
It is simple to apply a calibration in Origin.
Step 1. Given 2 datasets,
(1) Calibration Transmission Tc(Y) vs. Calibration Voltage Vc(X),and
(2) Measured Voltage Vm(Y) vs. time t(X),
store the cal set in columns A and B, and the measured set in C and D. Name them Vc, Tc, Vm, and t. Leave column designations to default (X,Y,Y,Y).
Step 2. Add a column and name it Tm for Measured Transmittance.
Step 3. Select the Tm column, right-click, and open the Set Column Values tool.
Step4. In the formula box, under "Col(E)=", enter
Col("Tc")(Col("Vm"))
Then click OK. This will generate linear-interpolated values of the Tc function for each of the Vm values measured.
Step5. Change the designation of the t column to X.
Step 6. Select columns t and Tm, and then graph as you would any XY dataset.
If you need to use more advanced interpolation, you can use the interp1 X-function (but I am not sure if X-functions were available in OriginPro 8.0). Please let me know if this helped.
DrBob |
| lkb0221 |
Posted - 01/23/2014 : 10:35:58 AM
Hi, Liana
You can use "Interpolate/Extrapolate Y from X" tool to implement this.
Please activate workbook and select "Analysis: Mathematics: Interpolate/Extrapolate Y from X" to open interp1 dialog.
Assuming you have T-V, V'-Tran two datasets and want to get T-Tran.
Use this tool to estimate values from V'-Tran to V-Tran, then your can plot T-Tran because they are corresponding to each other (linked by V)
Please click the question mark on the interp1 dialog to see more details.
Zheng
OriginLab |
|
|
