Conversation
…ethod in which the result was being converted to radians, which made no sense in the first place because eccentricity is a dimensionless measure of deviation from a circular orbit. The result of removing this improved some results and made some worse. The means by which this bug was found and the change was made is interesting, and I'll say more.
…tude was not being used. Again, in retrospect, it's unclear why I was not using it! A conversation with Cursor helped me worked through this. I made the relevant code and UT changes.
…te implementation
|
Hi @dbenn , please give me a couple of days. I’ll look at it over the weekend. |
There was a problem hiding this comment.
Hi @dbenn ,
Now the accuracy is quite good.
I took my observations of GSC_02743-01310 with time values converted from JD to HJD using AstroImageJ (see attachment: the first column is JD, the second is HJD converted by AstroImageJ)
Then I loaded the file to VStar (via Flexible Text plug-in) and converted the dates with the fixed plug-in. The absolute difference between VStar and AstroImageJ is less than 0.5s.
I think it's almost perfect.
Actually, the result could be even better because I rounded JD up to 5 digits after the point. In the original AstroImage file JD values have more digits (yet they are meaningless; they appear because the software I use -- SharpCap -- stores DATE-OBS with farctions of seconds).
|
Thanks @mpyat2! Although more testing could be done, it seems we're in agreement that this branch can be merged. |
2 participants
@mpyat2, I have introduced changes to fix this long-standing JD to HJD conversion problem.
@clkotnik, I wanted to include you as a reviewer here (even if only for your awareness) since we had interactions about this bug including a forum post you wrote, an embarrassingly long time ago. I got there in the end. I probably should have more strongly considered your suggestion at the time of using alternative Java code.
In the end, my robot overlords (Cursor) helped me. This use of GenAI and agentic AI in particular is a sweet spot I think. Please see issue #88 for commentary.