keep only one point per raster resolution value

[nmarieh]

hello! I have a dataset of points from which I'm extracting a number of values from 1km^2 resolution rasters. however for analytical purposes, I only want to keep one randomly selected point per 1km^2 cell. is there a way to filter my points and only select and keep one point per cell? thanks!

[DavidPike]

Why a randomly selected point within each cell, when the extracted value of any point within the cell would surely be the same?

[nmarieh]

sorry it's a bit confusing! since I'm using a bunch of different rasters to create this dataset (including one with a 30m res), I want to start from the outset with only one point per 1 km which is also the max coordinate uncertainty on my point localities, for ease of dataset management.

for example, two points might have identical precipitation values because they are within the same 1 km raster cell, or because they're very far apart and just happen to be in cells with the same precipitation value.

maybe there's a way to do it with the Generate Near Table tool that I haven't figured out? this is a pretty common need in species distribution modeling, so I feel like there must be a resource out there. 

thanks for your help!

[JayantaPoddar]

Check if these steps work (I haven't tried them though).

1. Execute Extract Values to Points to get the Raster values to points.

2. Keep a copy of the output Point Feature Class from last step (since next step will make changes to input Feature Class).

3. Execute Delete Identical tool. Choose the value and shape field. Define a higher value of XY tolerance (say 1000 meters).

[nmarieh]

hmm this didn't work for me, but I've almost got what I want by using Create Random Points and then setting the constraining feature class to be my existing dataset of points. the problem is I need to preset the number of points I want to end up, when really I just want ArcGIS to give me the max number of points possible such that no points are <1km apart.

[JayantaPoddar]

That is exactly what Delete Identical with predefined XY Tolerance was supposed to do.

What went wrong in these steps?

Considering the maximum distance of two points within a cell can be the "diagonal length" of the cell. So keep the XY Tolerance = 1414 Meters

Please share the issues in executing these steps, if any.

[nmarieh]

ahh gotcha, I think I've got it working now, if I set the number of points as the number in the current set and accept that I won't end up with that many. the first time I tried Delete Identical, I had the same number of points I'd started with. 

thanks for the help!