Refactored data_manager to avoid a problem arising when variables have different number of levels. Added test.py to run unit tests with coveraga report

Changed extracted levels for 6-hourly data. Now all variables will have the same levels to avoid concatenation errors

Changed order of variable renaming at send_file to avoid errors. Better management of valid_min and valid_max attributes

Created a class to encapsulate al the experiment management (get chunk list, get member str...) functionalities