Jump to content
GMS, SMS, and WMS User Forum
JKM

Scatter Data question

Recommended Posts

I have a series of five scatter sets that are rather large and it is taking forever to merge them. I may need to thin the data out some. Any suggestions?

Thanks,

 

John

 

Share this post


Link to post
Share on other sites

John,

SMS has tools to filter out scatter points, see the documentation below:

http://www.xmswiki.com/wiki/SMS:Scatter_Data_Menu#Scatter_Commands

You provide a filter angle which helps decide the extent of the filtering. I usually vary the angle a bit until I like what I am seeing for the remaining scatter set. So you may want to make a duplicate to preserve the original, if you do not end up liking what the filtering options give you.

I will add that I have run into large scatter sets where it also takes a long time to filter. These larger datasets end up chewing on the data while requiring a lot of RAM during the process. If the computer you are using doesn't have a significant amount of RAM, it will likely take a lot longer.

If these scatter sets were created from LiDAR tiles, there are other options for bringing them in that might work for you which handle the data more efficiently, processing wise.

Let us know how it goes.

-Cody

Edited by calberts

Share this post


Link to post
Share on other sites

Thanks Cody. I'll give that a try. I have 32G of Ram. What do you guy recommend for handling these large datasets?

Share this post


Link to post
Share on other sites

It depends on the situation, really. If I don't need to do a lot of changes to the TIN, I like to convert it to a raster format. High resolution rasters are handled very efficiently in SMS and with the realease of SMS 12.3, you can now use rasters for meshing and grid creation. Raw .LAS lidar files can be brought in as a raster or external software can be used to interpolate points to raster.

If the TIN data requires extensive edits and changes and I have to keep it in a TIN format, then I look for ways to cut the points down which could include filtering the data or deleting the outer extremities of the data which lies outside my project area.

Are any of those possibilities for your situation?

Share this post


Link to post
Share on other sites

What about utilizing multiple processor cores? Does SMS do that efficiently or is it better for the program to see the CPU as a single core? I'm considering upgrading my machine as I have several SMS models coming up with one in particular that will be quite large. Any hardware suggestions that you can make would be appreciated: Motherboard, Processor , RAM, etc.

Thanks, John

Share this post


Link to post
Share on other sites

John,

As far as some of the more routine or general processes go in SMS like filtering, retriangulating, and merging go, my understanding is that most only utilize a single core. And as far as running models goes in SMS, there are some numerical models which allow parallel processing such as ADCIRC, ADH, and even TUFLOW has a form of parallel processing with their GPU solver, but there are also many which do not have that capability. Are you planning on using any of those models? The general set of hardware requirements are listed at the bottom of the following page:

https://www.aquaveo.com/software/sms-surface-water-modeling-system-introduction

You could stick with the 32G of RAM that you have, and match that with a processor with the fastest clock speed that your budget will allow, as mentioned in those general set of requirements. 

-Cody

 

 

 

 

 

 

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×