Removed edge limit
making g++ happy
Added skeleton of edge stuff
Being more verbose about the SUSY scan parameters with which the algorithm was called
Provide possibility to do shape analysis for benchmark points as well (still testing)
Significantly sped up shape analysis by using asymptotic upper limits as direct input; not recomputing systematics over again for full cls; two separate shape droplets for asymptotic and final results (return asymptotic ones if no full cls is needed); introduced upper and lower cutoff within which full cls is required; using timeout script to make sure nothing funny is going on in combination tool; overriding standard tmp directory; forcing hadd to replace any existing file; using different subdirectory for full cls to avoid clashes; writing out timing; catching combine tool errors directly; using timeout to detect upper limit cutoff and avoid further conflicts; added final grid point to ensure safe computation; removed additional comma that would freak out bash; doing full cls computation for sms at every point; many more changes but i think space here is limited ...
Added scan identifier (t5zz,t5zzh,t5zzl etc.) based on scan directory; is read out when plotting scan results
fixed some uint vs int comparisons; commented out/removed unused variables; made Wall happy
Made error message for shapes clearer; replaced several cout's by dout's; made sure flipping is also done when falling back to CnC for limit setting
Catching shape limit exceptions and ignoring the limit if necessary; running both shape & CnC if shape analysis is requested to provide best possible limit
improved shape analysis implementation; now doing a first general guess using asymptotic limits, which is then refined; if the refined limit is within the exclusion reach, the full-blows cls limits are computed. also taking into account the signal integral (signal factor), and storing the raw limit (i.e. the limit obtained when normalizing to one)
Added a check for shapes to see if the efficiency is non-zero
Moved SUSY namespace, and using global msugra xsection file
Added new plot to show mll distribution as predicted and observed; expanded saved last configuration (for peak error in data)
Updated cross section saving algorithm to take into account filter efficiency
fixed a potential issue when running via crab that could have lead to aborts when transfers were unsucessful, therefore double checking successful transfer of required file(s)
Made scans more economic: files are on SE, point we want to consider gets downloaded, analyzed and the local copy is deleted; that way we significantly reduce the local storage required for the scans
Updated scan format for point-by-point files
Adapted susyscan for different types of scans. fixed a bug that could lead to a crash if too many samples were loaded; double bolted the scan type definition; made output slightly more verbose
Updated SUSYScan.C to allow quicker scans and also try out mSUGRA scans
Added some additional information for scans, such as the origin of the selected Z (interesting for mSUGRA)
Initial commit of Ice Cream versions
This form allows you to request diffs between any two revisions of this file. For each of the two "sides" of the diff, select a symbolic revision name using the selection box, or choose 'Use Text Field' and enter a numeric revision.