LabVIEW
Solve for XY plot
Solved!
An XY graph is a passive indicator, what matters is the data in the wire connected to it and you can analyze that any way you want.
I assume you want to find interpolated values for the lines connecting the points and this is trivial math. What have you tried? How well behaved are your points (e.g. are the x values sorted by point index? Do you have vertical segments? Can an x value contain more than one y point? etc.)
What is the datatype of your xy graph (1D complex array, array of points, cluster of x and y arrays, etc.)
If you have an array of points, interpolate 1D array would work best, so check the help for it.
Once you attach your code and some typical data, we can fix it for you in seconds.
See if this can give you some ideas.
(Probably needs to be adapted based on xy datatype, and other details)
Hey, thank you for the reply! We are trying to work with the cluster below. I have tried a couple things but no luck so far with my approach using property nodes.
Hi Robin,
@RobinWritesCode wrote:
What makes you think it is a good idea to label DBL floats as "String"?
Btw. you can easily convert your cluster of 1D array into the array of points used by Christian using one Unbundle and one Index&BundleClusterArray function. Then you can easily run Chistians code…
Hey,
I wasn't the one who labeled it this. It was generated by labview. There is no benefit to naming it a different type then it is.
Below is the code snippet as to the origin of the data.
Warmest regards,
Robin
Hi Robin,
@RobinWritesCode wrote:
Below is the code snippet as to the origin of the data.
Replace the Bundle node by Index&BundleClusterArray…
@RobinWritesCode wrote:
I wasn't the one who labeled it this. It was generated by labview. There is no benefit to naming it a different type then it is.
Because you started with string constants and didn't use BundleByName…
(You can also define the cluster with the Bundle node by wiring a typedefinition at the middle connector.)
The answer will be different if you want an approximation of the data as a whole or if you want piecewise interpolation between data points. If it's the former, you also need to know what form the relationship is in (i.e., is it linear, quadratic, exponential, etc.). If you want a more general fit but don't know the general form, you could use a spline fit.
So what does the data represent and what exactly are you trying to do with your interpolation?
(Also- you don't need the XY graph or property nodes to do this, you'd operate directly on the X and Y clusters- the thick yellow/gold wires going into the Bundle terminal).
Spline has the advantage that you don't need a real model, but has a disadvantage if the points have noise. What does your data represent? If you have a mathematical model for the function, a nonlinear fit would be much more robust.
If you use spline interpolation, make sure to use low-level functions and calculate the interpolant only once. In any case, you should skip linearly interpolated plot style and calculate the interpolated data as separate plot.
