twillis449 / ALBUS_ionosphere

software to determine ionosphere TEC and RM from GPS receiver data
Other
18 stars 9 forks source link

Change build to Python 2.7 #16

Closed bennahugo closed 1 year ago

bennahugo commented 1 year ago

Currently there is a yet to be found numerical issue when building under Python3.x which is not related to a dependency version change or division change as far as I can tell

There is also a parsing error where ALBUS under 3.6 rejects stations other than the first station on the stations list from what I've noticed from flipping between versions. It is possible a change in behaviour with iterators or another parsing issue which I am yet to trace down.

These are two distinct unrelated problems (I can reproduce the numerical issue with the single same station at Upington), however I've at least for the time being put my mind at ease that we get sensible stec solutions from the 2.7 build.

I will continue to dig on the #15 PR as I get time but for now I strongly suggest we swap the python 3.x build out for a 2.7 only build @twillis449 - it may be blind luck that this has not been seen elsewhere - it does not appear to have anything to do with the particular station.

@o-smirnov please in the interm redo your build to switch to Python 2.7. @mamkhari please deploy this version for your MSc. work on MeerKAT CAM as discussed

twillis449 commented 1 year ago

The issue between python 2.7 and python 3 has been resolved. There were some ASCII files that were being opened with the 'rb' flag with suggests that the files contain byte strings. This did not seem to actually affect the import of the files with python 2.7 but did affect the import of data with python 3. Changing the flags to just 'r' fixed this issue.

However, I am not able to replicate the issue concerning only getting data for one station on my computer.

bennahugo commented 1 year ago

I've made a new PR that fixes the latter @twillis449