DPlot

Hi,

I produce a lot of XYZ surfaces by exporting data from Excel using the plugin.

Instead of wrapping a surface over the top of the points, I'm trying to create separate surfaces passing through the mean and the median of the scatter points.

Is this possible?
Many thanks for your response,
Adrian Snell.

If you want those surfaces on the same plot: no, sorry. Currently you're limited to one surface per plot.

DPlotAdmin wrote:If you want those surfaces on the same plot: no, sorry. Currently you're limited to one surface per plot.

No I'd just want one surface per plot. If there is a way to get a mean or a median surface without writing a macro please let me know! I have searched!!

For the mean it's no problem: select the Z=f(X,Y) command on the Generate menu and use Z=$MEAN for the equation. That will of course be a featureless surface, just a flat plane. If all you want to do is show the mean value in a title line or note, just use $MEAN and it will be replaced by that value.

There's no way to do the same with the median, though. That's a good addition for the next release.

DPlotAdmin wrote:For the mean it's no problem: select the Z=f(X,Y) command on the Generate menu and use Z=$MEAN for the equation. That will of course be a featureless surface, just a flat plane. If all you want to do is show the mean value in a title line or note, just use $MEAN and it will be replaced by that value.

There's no way to do the same with the median, though. That's a good addition for the next release.

Thanks for the respose.

I have multiple data points at eaxh (X,Y) location. Instead of defining a surface at the max point at (x,y) (which DPlot currently does), I want to create a surface going through the mean of (x,y).

By the sounds of it this is not currently possible?

Many thanks.

Ah! Now I finally get it. No - not possible currently. I'll look into adding a switch to pick the mean, median, or (as is currently done) the max for duplicate X,Y values.

DPlotAdmin wrote:Ah! Now I finally get it. No - not possible currently. I'll look into adding a switch to pick the mean, median, or (as is currently done) the max for duplicate X,Y values.

Yes please! This will help me out no end!

Look forward to the update!

Many thanks,
Adrian.

After thinking about this some more, I'm not sure if a simple mean(x,y) surface would work.

Our points by nature are scattered, getting datapoints at mostly unique locations eg (10.001, 5.345), (10.002, 5.342).

In a 2D XY scatter plot we would typically overlay a moving average onto the scatter of points. What I'm thinking would work is to define a moving average at each point by taking the mean of the nearest x number of datapoints. So possibly have 2 separate options:

1. Simple mean and median surface options of multiple (x,y) datapoints
2. "moving average" mean surface by additionally defining the number of nearest surrounding points in (x,y,z) or to average over. Or take the mean at (x,y) then average the means in a predefined area surrounding (x,y) to get more of a moving average.