Friday, 22 July, 2022
Song of the Day: The Man Who Sold the World - Nirvana
Today I had the notion of trying to develop a star catalog for an earlier period of time which would correspond to several of the datasets that I am working with. Most of the ones at my disposal are several years old now, and so looking at the results in my analysis, I have developed a theory as to why when removing stars from the catalog, sometimes more images are capable of having their attitudes determined. Since they are older, the highest proper motion stars in the current catalog have moved relative to what they were expecting, which ruins any of the attitude determinations with those stars involved, until those stars are removed from the catalog.
Up until this point however, I have been unable to confirm this theory, as our recent catalogs are no longer compatible with the codebase, due to changes in how the binary files are formatted. Today, I discovered how the catalogs and inter-star-angle tables are generated, and attempted to create catalogs from past years.
Unfortunately, I am not seeing the results that I would expect to see. I also have no real way of confirming why the results are varying so wildly from what I would expect, as I personally have no way of even verifying that I have generated the catalogs correctly in the first place. Hopefully, next week I can reach out to one of the maintainers of that code and see how I might go about narrowing down some of the variables that might be causing me issues.