3D best fit tendline

[knfs]

Help much appreciated;

When using griddata and meshgrid to produce a 3D plot is there anyway to fit a trendline rather than just interpolate between the results.

Currently the experimental data used produces a somewhat 'wavey' plot. It would be much improved if a smoother plot could be achieved by fitting a surface through the data rather than interpolating between results.

Thanks

 

[rb1957]

i don't know how to do you are asking for, but ...

your data won't be improved by adding a trend line ...

the presentation of your data might, but then maybe the "waviness" is an important technical detail.

 

[knfs]

The experimental data i have shows some noise in the signal. A trendline would help to reduce the impact of this on the plot and allow a more 'useful' plot to be drawn.

comments welcomed

 

[touch84]

Report the data on Excel and go to the ‘Chart’ menu and click on the ‘Add Trendline’ option.
Click the ‘Options’ tab in the window . Check the ‘Display equation on chart’ option and click ‘OK’. This plots a linear fit and gives the slope and the offset of the fit.

 

[sreid]

Search the Web for "Data Smoothing." Mostly you'll find 2D info but these can be extended to work with surfaces.

 

[sugarshot]

Why not do the trendline in slices?

Smoothing data.. must be a NASA thing?

 

[sreid]

A couple of elementary data smoothing techniques are

Best Fit Straight Line (linear regression).

Moving Average Filter.