
<html><head><meta http-equiv="Content-Type" content="text/html; charset=us-ascii"></head><body style="word-wrap: break-word; -webkit-nbsp-mode: space; line-break: after-white-space;" class="">+1 Howard<br class=""><div><br class=""><blockquote type="cite" class=""><div class="">On Jul 18, 2019, at 11:34 AM, Norman Barker <<a href="mailto:norman.barker@gmail.com" class="">norman.barker@gmail.com</a>> wrote:</div><br class="Apple-interchange-newline"><div class=""><div dir="ltr" class="">Hi Even,<div class=""><br class=""></div><div class="">thanks for the clarification around indexing and nodata. This works great and when I have availability and your changes last in master I will upgrade the tiledb driver.</div><div class=""><br class=""></div><div class="">+1 from me.<br class=""></div><div class=""><br class=""></div><div class="">Norman</div></div><br class=""><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Thu, Jul 18, 2019 at 10:48 AM Even Rouault <<a href="mailto:even.rouault@spatialys.com" class="">even.rouault@spatialys.com</a>> wrote:<br class=""></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">Hi,<br class="">

<br class="">

> <br class="">

> I am in favour of the work on multi-dimensional arrays. There are two open<br class="">

> PRs <a href="https://github.com/rouault/gdal/pulls" rel="noreferrer" target="_blank" class="">https://github.com/rouault/gdal/pulls</a> with changes to the text. <br class="">

<br class="">

Oh I didn't see them before. I've merged the first one from Edzer. Answering <br class="">

yours now:<br class="">

<br class="">

> Though<br class="">

> I and my employer would like TileDB included in the list of formats (though<br class="">

> happy to wait until I have added the code to support this) <br class="">

<br class="">

I'd prefer to limit the scope of the RFC to the drivers I'll take care of. <br class="">

Other contributors are welcome to implement it for their own drivers of <br class="">

interest once this has landed into master.<br class="">

<br class="">

> the issues<br class="">

> around dimension ordering<br class="">

<br class="">

<a href="https://github.com/rouault/gdal/blob/rfc75/gdal/doc/source/user/" rel="noreferrer" target="_blank" class="">https://github.com/rouault/gdal/blob/rfc75/gdal/doc/source/user/</a><br class="">

multidim_raster_data_model.rst mentions<br class="">

<br class="">

"""Most drivers use the row-major convention for dimensions: that is, when <br class="">

considering that the array elemnts are stored consecutively in memory, the <br class="">

first dimension is the slowest varying one (in a 2D image, the row), and the <br class="">

last dimension the fastest varying one (in a 2D image, the colum). That <br class="">

convention is the default convention used for NumPy arrays, the MEM driver and <br class="">

the HDF5 and netCDF APIs. The GDAL API is mostly agnostic about that <br class="">

convention, except when passing a NULL array as the stride parameter for the <br class="">

:cpp:func:`GDALAbstractMDArray::Read` and <br class="">

:cpp:func:`GDALAbstractMDArray::Write` methods. You can refer to NumPy <br class="">

documentation about multidimensional array indeing order issues:<br class="">

<a href="https://docs.scipy.org/doc/numpy/reference/internals.html#multidimensional-array-indexing-order-issues" rel="noreferrer" target="_blank" class="">https://docs.scipy.org/doc/numpy/reference/internals.html#multidimensional-array-indexing-order-issues</a> """<br class="">

<br class="">

Does that answer your concern / question ?<br class="">

<br class="">

> and defining NULL are two things I would like to<br class="">

> discuss.<br class="">

<br class="">

Yes, nodata support is there. Mentionned in<br class="">

<a href="https://github.com/rouault/gdal/blob/rfc75/gdal/doc/source/user/" rel="noreferrer" target="_blank" class="">https://github.com/rouault/gdal/blob/rfc75/gdal/doc/source/user/</a><br class="">

multidim_raster_data_model.rst#multidimensional-array<br class="">

<br class="">

and in the C++ API:<br class="">

<a href="https://github.com/rouault/gdal/blob/rfc75/gdal/gcore/gdal_priv.h#L2257" rel="noreferrer" target="_blank" class="">https://github.com/rouault/gdal/blob/rfc75/gdal/gcore/gdal_priv.h#L2257</a><br class="">

<br class="">

so 2 virtual methods that can be implemented by drivers:<br class="">

const void* GetRawNoDataValue() const;<br class="">

bool SetRawNoDataValue(const void* pRawNoData);<br class="">

<br class="">

and 2 helpers that get/set as double for the common cases.<br class="">

<br class="">

Even<br class="">

<br class="">

-- <br class="">

Spatialys - Geospatial professional services<br class="">

<a href="http://www.spatialys.com/" rel="noreferrer" target="_blank" class="">http://www.spatialys.com</a><br class="">

</blockquote></div>

_______________________________________________<br class="">gdal-dev mailing list<br class=""><a href="mailto:gdal-dev@lists.osgeo.org" class="">gdal-dev@lists.osgeo.org</a><br class="">https://lists.osgeo.org/mailman/listinfo/gdal-dev</div></blockquote></div><br class=""></body></html>