Avoid issues with ListPolarPlot and ListSurfacePlot3D?

[juergenkanz (Jürgen Kanz)]

Hi,
I am trying to create a nice ListPolarPlot and ListSurfacePlot3D. Unfortunately the results are not as expected.

Examples:
1.Problem: I do not get a full ListpolarPlot, which means I am loosing PolarTicks and the entire polar circle.

2. Problem: I do not get a full ListSurfacePlot3D. Data in the plot are cutted so that the entire appearance of the “body” gets wrong.

Attached are some files to play with. The “Graphics-Community.nb” creates the mentioned plots based on some real world data.

Question: What can I do to improve the plots?
Thanks for your help in advance.

Graphics-Community.nb (537 KB)
XYZ.mx (106 KB)
pairsC0C180.mx (1.31 KB)
pairsC90C270.mx (1.31 KB)

[kmartin (Kyle Martin)]

1. ListPolarPlot seems to choose a bad PlotRange in this case, particularly with the Interpolation option. You can workaround this with the PlotRange and PlotRangePadding options, i.e.

ListPolarPlot[{pairsC0C180, pairsC90C270}, PlotStyle -> {Red, Blue}, 
 Joined -> True, InterpolationOrder -> 2, PolarAxes -> True, 
 Joined -> True, PolarGridLines -> True, 
 PolarTicks -> {"Degrees", Automatic}, 
 PlotLabel -> HoldForm["LID in Polar Coordinates [cd/klm]"], 
 LabelStyle -> {16, GrayLevel[0]}, AxesStyle -> Gray, 
 AspectRatio -> Automatic, ImageSize -> Large, 
 PlotRange -> All, PlotRangePadding -> Scaled[.08]]

I’ll report this issue with our visualization team so that we can look to improve the functionality in a future release.

2. The ListSurfacePlot3D is trickier. Is it possible to obtain more points from your source data?

[juergenkanz (Jürgen Kanz)]

Hi Kyle,

Thank you very much for the problem #1 workaround.

Regarding your question: the number of points depends on the measurement grid, hence to add interpolated points wouldn’t help.
Instead I have created additional pair and xyz files of typical distributions.

…no2.mx files belong to a axially symmetrical distribution. Result is okay.

…no3.mx files belong to a different asymmetrical distribution compared to the first example. Result is still okay.

…no4.mx files represent other symmetries. Please compare the meaningful ListPointPlot3D with the ListSurfacePlot3D. At least I have problems to recognize the “body”.

You can still use above mentioned notebook to visualize the files.

Thank you for support.

pairsC0C180no2.mx (3.62 KB)
pairsC0C180no3.mx (17.7 KB)
pairsC0C180no4.mx (2.29 KB)
pairsC90C270no2.mx (3.62 KB)
pairsC90C270no3.mx (17.6 KB)

[juergenkanz (Jürgen Kanz)]

Upload Part 2:

XYZno2.mx (68.9 KB)
XYZno3.mx (1.31 MB)
XYZno4.mx (172 KB)
pairsC90C270no4.mx (2.32 KB)

[juergenkanz (Jürgen Kanz)]

Hi Kyle, I am sorry to say that the workaround doesn’t work in a sustainable manner. Several input line executions or notebook evaluations lead to a “back and forth” of graphic sizes.