Hi,
This question is a combined question for McIDAS User Services and HP. I am planning on transitioning some of my scripts which for lack of a good reason, use my own rgb formulae rather than the EUMETSAT plugin. In particular, I would like to transition an airmass rgb script. In the EUMETSAT RGB version, the bands are subtracted directly (ch05-ch06). I am wondering if there is any advantage to using sub(ch05-ch06) or if these are equivalent? If I am loading data from individual netCDF files, there could be navigation differences and I was hoping that sub() would check the navigation first. Next, I had also added a mask in my files in case a file had missing data erroneously set to zero. Is this being overly cautious with H8 ADDE data?
I am just curious, I think it is quite possible that adding sub() and mask() may be extra work, but I wanted to check before transitioning.
Thanks,
Joleen
EUMETSAT Airmass RGB question
Re: EUMETSAT Airmass RGB question
Hi Joleen-
Jon found the code that sub ultimately calls before handing things off to the same VisAD code that handles “band1 - band2”:
https://github.com/mcidasv/mcidasv/blob ... #L300-L359
The javadoc contains the following:
"This method handles making the data as compatible as possible before actually going off and doing the math."
And that makes sub seem like the better choice due to your concerns about navigation differences. With that said, Bob ran some simple tests of sub(a,b) and a-b with imagery and local grids. The results were identical, but those were with examples with known identical domains.
Hope this information helps.
Thanks, Jay
Jon found the code that sub ultimately calls before handing things off to the same VisAD code that handles “band1 - band2”:
https://github.com/mcidasv/mcidasv/blob ... #L300-L359
The javadoc contains the following:
"This method handles making the data as compatible as possible before actually going off and doing the math."
And that makes sub seem like the better choice due to your concerns about navigation differences. With that said, Bob ran some simple tests of sub(a,b) and a-b with imagery and local grids. The results were identical, but those were with examples with known identical domains.
Hope this information helps.
Thanks, Jay