fixed bug in DldProcessorV2.py, now it actually runs, even if the number of partitions is not divisible by 8. However, for extreme binning it is still inefficient.

Merge branch 'master' of gitlab.phys.ethz.ch:eduphys/OfflineAnalysis

added a new pair of classes: The DldProcessor is used to analyze the dask dataframes later, reading them from a hdf5 file. The DldFlashDataframeCreator is derived from the DldProcessor and can read a run. It is also used to store the run in a hdf5 file containing the dask dataframes. This way, we just need to distribute the dask dataframes as well as the DldProcessor so the user can analyze it without all the FLASH specific libraries.

Removed compiled python file

Tests on the Jupyter Notebook file

DldFlashProcessor. Much more convenient, no double code :-)

Test for HDF5 files (alpha)

PDF from taifun elog to compare photoemission spectra

IPython Notebook for Auswertung

Without hdf5 files