reportCEScalib was not working when starting it with output.R

Plot supplycurves for all years not just one year.

* Enable prepare_and_run.R to restart fixed runs (https://github.com/remindmodel/remind/issues/112)
* Enable start.R to restart non-optimal runs

Change the reference scenario (for calculating the BECCS bound, see previous commit) from input_ref.gdx to the values of the own previous nash iteration.

EDGE-T scenarios update; EDGE-T supports non-fuel divided by model type and differentiated learning.

change version of gdx being used for NDC target calculation (now input_bau.gdx) so that correct NDC targets are also calculated for non-SSP2 pathways

- added documentation to the output script.
- updated the multi-comparison-matrix.
- updated the prefix used for coupled runs.
- correctly handle the number of cores to be used in parallel.

Add .Rprofile to the list of files that are copied into the results folder *before* prepare_and_run is called.

gdx neutral: clean scenarios in scenario_config_coupled_SSPSDP.csv and compareScenarios.R, add start and end time (of the job not of GAMS) to runstatistics.