Python 2 --> 3 transition plan

[abostroem]

Preliminary work:

• Py2-->3 script on a new branch
• Update docker
• look at diff and check before committing
• figure out cosmic software - consider switching to astroscrappy. From Stefano updating the DLT40 pipeline: there is a new version of lacosmic that is python 3 compatible, however it is quite different and the input parameter are a bit different while input parameter for astroscrappy are the same that we use
• What does the pipeline use to interact with the database?
• Trial and error run and fix

Testing plan:

• Step 1: get a file to run all the way through in Python 2 and Python 3 and compare final photometry (0.01 mag diff acceptable)
• Step 2: Choose a 100 frames known supernova with good reduced data and reduce it myself for consistenty
  Check that everything is within 0.01 mag (1% error)
• Step 3: Test different configurations, paths through the software, intermediate files

Check:

• database
• catalogs
• sn2 files
• all intermediate fits files

Processes to check:

• LCOGTingest, ingesttar
• wcs, psf, psfmag, zcat, mag
• checkwcs, checkpsf, checkmag, getmag
• Diff imaging with LCO templates (hotpants and pyzogy)
  • -s diff, template, cosmic
• Diff imaging sloan

run with -i and --show and --no_iraf

Choose one dataset from each of the following:

• first letter chips, second letter is for camera
• fairchild chips and spectral camera (fs)
• fairchip chip and lco camera (fl)
• fairchild chips and arcon controler (fa)
• kodak chips in an sbig camera 1m and 0.4m (kb)
• sony chips in qhy camera (sq)
• e2v chips and Princeton instrument cameras (ep)