save_bin_data == False breaks lightcurve creation

[pivosb]

I've been having unexpected issues with the lightcurve generation in fermipy once I set save_bin_data to False.

Here's the error:

---------------------------------------------------------------------------
StopIteration                             Traceback (most recent call last)
/tmp/ipykernel_134231/3996349889.py in <module>
      1 # Generate a lightcurve with 1-day binning
      2 #lc = gta.lightcurve('1es0229+200', binsz=86400. * 90)
----> 3 lc = gta.lightcurve('1es0229+200', time_bins=[611334005, 628423205, 645512405])

/afs/ifh.de/group/cta/scratch/pedroivo/Software/anaconda3/envs/fermi/lib/python3.7/site-packages/fermipy/lightcurve.py in lightcurve(self, name, **kwargs)
    287         self.logger.info('Computing Lightcurve for %s' % name)
    288 
--> 289         o = self._make_lc(name, **config)
    290         filename = utils.format_filename(self.workdir, 'lightcurve',
    291                                          prefix=[config['prefix'],

/afs/ifh.de/group/cta/scratch/pedroivo/Software/anaconda3/envs/fermi/lib/python3.7/site-packages/fermipy/lightcurve.py in _make_lc(self, name, **kwargs)
    451         for i, time in enumerate(zip(times[:-1], times[1:])):
    452             print(i, time)
--> 453             next_fit = next(mapo)
    454 
    455             if not 'fit_success' in next_fit:

StopIteration:  

After some digging, I realized that if save_bin_data is set to False, then in

fermipy/fermipy/lightcurve.py

Lines 440 to 442 in b016308

	if not kwargs.get('save_bin_data', False):
	for m in mapo:
	shutil.rmtree(m['config']['fileio']['outdir'])

you iterate over mapo , and then in

fermipy/fermipy/lightcurve.py

Line 450 in b016308

next_fit = next(mapo)

the iterator is exhausted and raises the StopIteration error.

The solution I found for now is to simply set save_bin_data to True again, and it works just fine.

[henrikef]

I see what you mean. The iterator doesn't reset after you loop over it the first time. I think we can fix it by moving the file deletion into the 2nd loop. Let me give it a try.

[henrikef]

Take a look at #474 (branch lc_fixes) if you'd like.