[pdal] write an in-memory array of points to a writer ... using Python bindings

Michael Rosen michael.rosen at gmail.com
Thu Apr 20 08:59:37 PDT 2017


Thanks for the explanation. One  alternative might be to add an HDF reader
/ writer.   Thoughts on the feasibility of this?  Thoughts on using the
official HDF library vs leveraging the implementation in GDAL (which I
assume is different) ?
On Thu, Apr 20, 2017 at 7:52 AM Howard Butler <howard at hobu.co> wrote:

>
> On Apr 19, 2017, at 6:29 PM, Michael Rosen <michael.rosen at gmail.com>
> wrote:
>
> Can we use the Python Bindings to create a LAS file from values we hold in
> Memory?  In my case, I'm using the HDF Python library to read a .h5 file
> and need to convert the results to LAS.
>
> In the Writing with PDAL documentation
> <https://www.pdal.io/tutorial/writing.html#writing>, Bradley Chambers
> shows how to do this in C++ ... but it looks like we're missing some
> machinery to do it via Python.
>
> In the test for the Python Bindings
> <https://github.com/PDAL/PDAL/blob/master/python/test/test_pipeline.py>,
> we see how to write the output from a filter to an array ... but not the
> other way around.
>
>
> Not at this time, but it's been on my queue for Python stuff. I filed a
> ticket at https://github.com/PDAL/PDAL/issues/1573 on the topic.
>
> Part of the reason it isn't done yet is I haven't been sure how it would
> work. There's a few challenges
>
> - We aren't going to do IPC, so the only way it could be used is in Python
> extension scenarios https://pypi.python.org/pypi/pdal .
> - PDAL has explicitly named dimensions https://www.pdal.io/dimensions.html and
> Numpy has array names. The user is going to have to map their arrays to
> known PDAL names or things will be quite mushy, especially when translating
> to other formats.
> - Unsure what to do about multiple arrays at once, but I presume we'll
> just disallow it for starters
> - Metadata, SRS, and  construction and assignment
>
> None of this is too difficult, but there's still a significant effort
> required to complete it.
>
> Howard
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/pdal/attachments/20170420/be59a39b/attachment.html>


More information about the pdal mailing list