Support stats type NanCount

[pford]

For CyberSKA requirement, no corresponding stat in casacore.

[veggiesaurus]

@arusstaylor can you comment on whether this is essential? Haven't seen a NaN count statistic in viewers other than CyberSKA before.

[arusstaylor]

I don’t think this is essential as long as the pixel count does not include Nans. Russ

…

Sent from my iPad

On 05 Apr 2019, at 07:41, Angus Comrie ***@***.***> wrote: @arusstaylor can you comment on whether this is essential? Haven't seen a NaN count statistic in viewers other than CyberSKA before. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread.

[veggiesaurus]

@pford I believe the cassacore pixel count excludes NaNs at this point. Should we close this issue then?

[arusstaylor]

Fine with me to close. Russ

…

On 05 Apr 2019, at 09:55, Angus Comrie ***@***.***> wrote: @pford <https://github.com/pford> I believe the cassacore pixel count excludes NaNs at this point. Should we close this issue then? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#161 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AIotLWBeRBLVifLmmvX1fDZZz9xAOkzgks5vdwF2gaJpZM4cc9Ps>.

[pford]

Yes, NumPixels excludes NaNs. Will close. This CARTA::StatType should be removed from the ICD.

[pford]

Needed for release 1.2 based on testing feedback.

[pford]

As I understand it, this is extremely difficult in casacore. Images are masked so that unmasked pixels may be used for calculations, which means that for a region, a pixel may be masked because it is NaN, it is outside the region, or both. Therefore impossible to count "in the region but value is NaN".

[veggiesaurus]

@pford I guess we can icebox this for now then