<div dir="ltr"><div>Thanks Doug.. <br><br></div><div>My recollection is that these items were originally going to be added as an image-specific extension of ObsCore.<br></div><div>This document has them incorporated into the ObsCore table proper, and the language (number of samples)<br></div><div>seems generalized to include non-pixelated datasets.<br><br></div><div>Since these are being added to the required element list, the text should be very clear about what goes in there for unbinned datasets.<br><br></div>Mark<br><br></div><div class="gmail_extra"><br><div class="gmail_quote">On Wed, Jun 24, 2015 at 9:06 PM, Douglas Tody <span dir="ltr"><<a href="mailto:dtody@nrao.edu" target="_blank">dtody@nrao.edu</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><span class="">On Wed, 24 Jun 2015, CresitelloDittmar, Mark wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
pg 25: Axes dimensions<br>
Outside of the pixelated image context, what values do these hold?<br>
For example, the sparse cube (ie: event list). do each of these hold<br>
the #events? or are s_dim1,s_dim2 == the span along that axis? and<br>
if so, in what frame?<br>
It seems that #events is closest to the description as the number of<br>
samples.<br>
</blockquote>
<br></span>
For event data which is not imaged, such as an event list, probably the<br>
best one can do is have t_dim specify the number of events. The spatial<br>
coverage would be specified with s_fov and the spectral coverage with<br>
em_min/max. The binning on the spatial and spectral axes is meaningless<br>
and should be zero. If a precomputed image for the observation is also<br>
provided, that would have all the dim values and would share the same<br>
obs_id.<br>
<br>
- Doug<br>
</blockquote></div><br></div>