Bring the implementation which already exists from the wavelet to the master branch. Adapt the code for line in time multi threading to images.

• make sure the git repo has everything from the SVN
• make sure code still works - especially all tests

We don't want that PynPoint directly prints to the Terminal.

• logfile
• plots for the current data
• plot graph of the current architecture

Find out which parts of PynPoint already have test cases. Make create Issues for missing test cases with priorities!

The get_all() function reads the data as float64 even if it is only float32. Change this or add a parameter and fix unit tests.

1 solution:

• turn them on in a debug mode (config file)

2 solution:

• read only a subset (all needed) header keys
• use the dict mentioned in a "Easy to use" issue

Details need to be discussed.

Silvan send me his code, I will add this functionality.

I wanted to install PynPoint on Linux Ubunt 16.04 LTS and couldn't import it due to the an ImportError. I proceeded as follows:

• Cloned the repo from GitHub

• Run (Python 2.7) python2 setup.py install, which run successfully

• When importing PynPoint there was an ImportError due to the numba package

     OSError: /usr/local/lib/python2.7/dist-packages/llvmlite-0.14.0-py2.7.egg/llvmlite/binding/libllvmlite.so: cannot open shared object file: No such file or directory 

• Same error occured when trying to import numba

Solution
It was suggested here that one should use conda to handle the environments. I downloaded Anaconda and now I can import both numba and PynPoint without any errors.

We need to make sure that our tests are being run in a CI mode

We need to find out how to move to a public git repo. Important is especially how we can protect the master branch.

Also look which modules might be changed in order to benefit from this!

Some Header keywords have different names in different datasets but contain the same information. Create a central dict like:
#Header Key# = a global key in the database

Change all modules (reading and processing) which have hard coded Header keywords.

Write a style guide for PynPoint:

• names of members
• pylint
• short .pdf

We need to discuss a way to increase the impact of the new PynPoint release.

• Paper
• Conference
• Tutorials
• Twitter / Facebook ...

We want to write a news Letter to all people using PynPoint at the moment to present an overview of the new features.