Data Cull

[Lurking]

Just a thought... and if it's in there already, my apologies.

How about a way to cull all data outside of the the bounds of what your extents are set to? I ask because after grinding through a massive terrain set, I only wanted to keep that part that I was interested in.

My current method is to take the 129Meg+ SRTM .asc file and convert it to a generic point text file [with LandSerf] after paring it down to the area of interest, then replacing the tab delimiter with a comma [with UltraEdit], then into Dplot with the terrain data. I still get sizable files and after tweeking it down to what I'm interested in, and I'd like to toss all the data outside of the area of interest.

[DPlotAdmin]

While viewing in 2D, see Edit>Crop. When looking at the entire data set it doesn't give the sort of control you probably want, but if you've already entered the extents you'd like then it is handy: just drag the cursor from one corner of the plot to the other.

[Lurking]

Thanks for the response. Handy for 2D but I sort of need that 3rd axis data. My version of Excel can't handle more than 65K points, and to down-sample it prior to the importing costs resolution.

Still mulling this over.

[DPlotAdmin]

I haven't previously seen a need for a "Clip Z" command. It would be easy enough to add, but you should be aware that without also setting the plot extents to those same Z values you'll still get triangles generated where you expect to see holes.

[Lurking]

Mainly I'm looking to use it on scatter plots. I ran into the weird triangle effects when deleting parts of a surface plot. Most un-nerving.

[DPlotAdmin]

Just an FYI: It really isn't that weird once you consider what you're giving DPlot to work with (X,Y,Z points). There's no information in that data about where boundaries should be. DPlot generates a convex triangular mesh, and for X,Y,Z points with no boundary conditions specified there's really no solution that makes better sense. You can specify X,Y boundary points with Options>Define Boundary or specify the extents with Options>Extents/Intervals/Size, but in either case that information will not be saved to a simple CSV file, which it sounds like is what you're after.

[Lurking]

I have found the ability to get partway there... at least it's workable. Not elegant, but highly functional. I was able to wrangle UltraEdit into a sort by column and could slice off the ends of the data I didn't want.


The terrain was pared down to the area of interest. The lower plot was done separately and pasted onto the bottom.





Interesting... It labeled the Deg North as East... and the West axis is correct. The scatter plot of the quakes came out correct.

Here is a merged version of the quake scatter and the terrain surface. Now that is weird. Dunno why it did that.

[Lurking]

Okay... this is making my head hurt. I make a scatter of the surface and the N/S/E/W comes out correct. I make a surface of the scatter, still good. Scatter of the surface again, still good. Kill all of those and go back to the original, take away the N/S/E/W, switch to default number format. Switch back to deg/min, turn on the N/S/E/W and it's wrong.

Hmm... You don't think that there could be an extra line feed or character horking my data set do you?

[DPlotAdmin]

Sorry for the trouble. You found a bug. It's fixed here and will be available later today.

[Lurking]

Actually... (to me) that's a good thing. I thought I had mangled part of the data.

[DPlotAdmin]

There's no "Clip Z" - I'll try to have that in the next release. But the bug you found is fixed. If you signed up for the mailing list you'll soon get an e-mail with download instructions. Otherwise you can use "Check for Updates" on the Help menu.

And thanks for pointing out this problem.

[Lurking]

Works like a champ. Thank You!

As for the cull, it's not a show stopper, I just didn't know If I had missed it somewhere.