Repository navigation
BUG: sw_f107 historic data not available after 2018-04-30 #11
Description
Description
The historic data does not download past April 2018. This appears to be an issue on the LISIRD side of things.
A new data source should be found, or at minimum it should be documented in the docstrings.
To Reproduce this bug:
import GeospaceDataManagement as gdm
import datetime as dt
solar = gdm.Instrument(inst_module=gdm.instruments.sw_f107, tag='historic')
solar.download(dt.datetime(2018, 4, 1), dt.datetime(2018, 6, 1))
solar.files.files
The code runs successfully, but only April data is downloaded. This appears to be on the lisird side of things, as constructing the url string manually produces the same result.
Test configuration
- OS: Linux Ubuntu Questing Quokka
- Python 3.14.7
Additional context
n/a
Activity
aburrell commentedon Oct 21, 2026
Ah yes, this is the last date they support. I currently don't have anything hard coded (like the start date), in case they increase their coverage in the future.
What kind of solution are you looking for?
jklenzing commentedon Oct 21, 2026
Probably a note in the docstring. Is there another instrument currently that supports later dates? Or is this a gap in current support for historical data?
aburrell commentedon Oct 22, 2026
Use the 'prelim' data. This is why I have the 'combine_f107' function, there is no place to get the entire F10.7 data set.
Not to be argumentative, but I am not wild about a note in the docstring. None of the docstrings really make a note about the period of time the external sites have data coverage. And since this could change at any time (either to be more or less), it seems like a nightmare to try and maintain.
I'm guessing the thing you actually found annoying was that the download didn't inform you that only part of the data was downloaded. What about something that reports on the success/failure of the requested download period (not data set specific)?
Also, check out
docs/examples/ex_f107.rstfor a bit of code to load a continuous swath of F10.7 from multiple sources.jklenzing commentedon Oct 23, 2026
That sounds like a good improvement. Largely the issue was the code did not inform me about the data gap. This solution has the advantage that it does not need to be updated if the host changes what they support.
aburrell commentedon Oct 23, 2026
I added a logger warning to the existing updates in #10. Feel free to use that branch until the updates go through.