Closed orionlee closed 1 year ago
The list in the dialog UI was not being populated early. Fixed that.
There are names appearing by default that should not be since each source is different, e.g. obscode will appear in AID data but not in TESS.
This requires some thought to implement properly. Working through a few approaches. In principle should be simple, but less so than I'd like.
@dbenn Maybe I misunderstand how the issue is addressed. But when I run the latest VStar on master
(d2d2317d) , and open a TESS SPOC or QLP fits file, and use View > VeLa Filter
quality
(specific to TESS) is not in the Observation Properties dropdownIn contrast, the drop down for regular expression filter reflects the actual available columns:
Thanks @orionlee. Investigating.
@orionlee: as per the new pull request, I've fixed the arbitrary column names case. I've also removed obscode as a special case. There are other names than columns no matter what the data source, e.g. as a convenience for time (t, jd, ...), magnitude (mag), and band (shortband, e.g. V vs Johnson V), series (can be the same as band, but not always, specifically when a user defined series is created, e.g. from a filter). Some names have underscores added for VeLa but are mapped back to the underlying name.
It'd be great if there is some way to know the names for the columns to be used in a VeLa filter.
Right now one can find out some of them from the examples in user manual or wiki, but one cannot know easily if a column can be used in VeLa filter, and if so, the valid spelling. E.g., can one filter on calendar date?
It might be a headache to document them, given they are source specific.
One possibility is that VStar UI provides a list of valid names for the current source.