months problem
Closed this issue · 9 comments
I installed RMextract. But I've encountered some really weird problem - It seems unable to generate codg file names if the month being specified is 4 (April) or less. e,g
following does work for month >= 5, fails for month <= 4
START_TIME="2016/4/09 00:00:50.1234" # fails
END_TIME="2016/4/09 01:00:50.456" # fails
but following works
START_TIME="2016/5/09 00:00:50.1234" # works
END_TIME="2016/5/09 01:00:50.456" # works
See the test script
i'm using ubuntu 16.04 linux
I get the same errors on an ubuntu 18.04 system
BTW: Its also a bit annoying that all those 'DEBUG:root ...' messages are presently turned on by default.
Strange indeed, I cannot reproduce this. Could it be that the files on the server are corrupted, or that you have a corrupted IONEX file stored somewhere? It would be helpful if you could give the reported error.
I will put in some additional debug statements. I see that you have switched to using ftplib instead of the older (and simpler to me!) pycurl interface. So your coding has become somewhat harder for me to understand. I don't think this has anything to do with corrupted IONEX files because at least then the system should download something which would cause a local crash when trying to read and extract data from the file on my machine, What's happening is that it just reports that it could not find any 'codg' file to download. I'll get back to you with some more detailed output.
I switched to ftplib to make the search path more general, such that if the server decides to change the path there is still a fair change to find the files. The problem is that I cannot reproduce the issue.
I have an old hard drive here which I will use to do a completely new install of ubuntu 18,04 wiping all previous stuff of the disk. Then I'll see what I get.
OK - found I already had a bare-bones install of ubuntu 18.04. After installing all supporting packages I ran the agw_examples.py script but changed the final start and end times to
START_TIME="2014/1/09 00:00:00"
END_TIME="2014/1/09 23:59:59"
I then get:
llis@twillis-Latitude-E6420:~/Downloads/RMextract-0.3/examples$ python agw_examples.py
We will need PyEphem to perform calculations!
the accuracy of results might decease a bit
PyEphem will be used to perform calculations!
observing in direction RA, DEC in radians 5.15199451295 -0.349065850399
pointing [ 5.15199451 -0.34906585]
('ephem start and end time ', 56666.0, 56666.99998842599)
INFO:root:Retrieving data for 2014 or 14008
INFO:root:Retrieving data from gnss/products/ionex/2014
INFO:root:
Traceback (most recent call last):
File "agw_examples.py", line 64, in
result = getRM(use_azel=use_azel,radec=pointing,out_file='test_report',object=OBJECT,start_time=START_TIME,end_time=END_TIME, timestep=TIME_STEP,stat_positions= upington_position,useEMM=True,TIME_OFFSET=TIME_OFFSET)
File "/usr/local/lib/python2.7/dist-packages/RMextract-0.1-py2.7-linux-x86_64.egg/RMextract/getRM.py", line 153, in getRM
ionexf=ionex.getIONEXfile(time=date_parms,server=server,prefix=prefix,outpath=ionexPath)
File "/usr/local/lib/python2.7/dist-packages/RMextract-0.1-py2.7-linux-x86_64.egg/RMextract/getIONEX.py", line 544, in getIONEXfile
return _get_IONEX_file(time, server, prefix, outpath, overwrite)
File "/usr/local/lib/python2.7/dist-packages/RMextract-0.1-py2.7-linux-x86_64.egg/RMextract/getIONEX.py", line 507, in _get_IONEX_file
prefix)
AssertionError: No files found on ftp://cddis.gsfc.nasa.gov/gnss/products/ionex/ for codg
THank you. THe before after April 9 issue must be related to the dayofyear going from 99 to 100. I see that I have put in a possibly related fix about a month ago. Could you check out the latest code directly from master (I also changed logging level to ERROR in that same fix, so it seems you do not have it). I created a V0.4 for you as well.
Yea team! Looks like you've fixed things with v 0.4. :)
The only odd thing I've noticed is that you generate an error message
/usr/local/lib/python2.7/dist-packages/RMextract-0.1-py2.7-linux-x86_64.egg/RMextract/getIONEX.py:188: RuntimeWarning: invalid value encountered in divide
weights = weights / _steps
I'm not sure what you are doing there but it does not seem to affect the outcome.
Closing this issue.