[pdal] How to read a text file

[Howard Butler]

> On Mar 24, 2017, at 2:34 PM, Michael Rosen <michael.rosen at gmail.com> wrote:
> 
> Thanks guys.
> 
> 
> So, I'm looking at the Sithole-Vosselman "Filter Experiment" paper from 2004.  The data they provide is in text format and sadly, puts two points on a line (first and last return, though they don't give return number nor total number of returns).
> 
> data: http://www.itc.nl/isprswgIII-3/filtertest/
> 
> [X1 - First Return] - x coordinate (Easting)
> [Y1 - First Return] - y coordinate (Northing)
> [Z1 - First Return] - z coordinate (Height)
> [I1 - First Return] - Intensity of returned pulse
> [X2 - Last  Return] - x coordinate (Easting)
> [Y2 - Last  Return] - y coordinate (Northing)
> [Z2 - Last  Return] - z coordinate (Height)
> [I2 - Last  Return] - Intensity of returned pulse
> 
> 
> To move forward I can, of course, preprocess the file so that these read in as two separate, unrelated, points.  However, a robustness test for PDAL would be if I could make a pipeline that would read it in as presented.  Could someone comment on how feasible that is.

I think this is possible with filters.programmable and Python, but there might be a gotcha or two. Assuming you fix the header thing and you read in the points, you should be able to make new points for each of the second points in the line using numpy.