===============
Nominal result:
===============

1) results_prefit.csv
2) results_postfit.csv

The two .csv files contain the nominal result: 
1) provides the pre-fit SM background estimates;
2) provides the post-fit SM background estimates (in which the fit is
performed to data in the signal region and all control regions).  

The covariance and correlation matrices are not provided.

The first line of each .csv file defines the contents of the columns:
njet,btag,htlow,hthigh,mhtlow,mhthigh,bkg,bkgErr,data

Notes: 
- the first 6 columns define the bin in terms of njet, nb, HT, and HTmiss
- bkgdErr contains both stat. and syst. contributions.
- the njet and nb categories are identified through 4-character codes:
  - eqNj: exactly 'N' jets (symm)
  - eqNa: exactly 'N' jets (asym)
  - eqNb: exactly 'N' b-tagged jets
  - geNb: at least 'N' b-tagged jets
- "symm": implies symmetric pT thresholds of >100GeV on the two leading jets
- "asym": implies asymmetric pT thresholds of >100 and >40GeV on the two leading jets

============================================
Result based on a simplified binning scheme:
============================================

This result is based on a simplified categorisation of candidate
signal events. The SM background estimates are obtained from a
"CR-only fit", in which a simultaneous fit to data yields in all
control regions is performed, while the signal region is excluded
from the fit.

Event yields are provided for eight representative "topologies",
defined in terms of requirements on Njet and Nb. For each topology,
the event yields are integrated over the full HT range, HT > 200 GeV,
and categorised according to eight bins in HTmiss. This categorisation
scheme leads to 64 bins that are exclusive, contiguous, and provide
complete coverage of the signal region. The corresponding background
estimates are determined from the nominal likelihood model. The
covariance matrix for the background estimates in the 64 bins is also
available. This minimal set of information can be used to aid the
reinterpretation of the search result using physics models not covered
in this paper. The procedure using a simplified likelihood approach,
as described in Ref. [1] below, demonstrates a good degree of accuracy
for a wide range of models. A discussion of the assumptions and
approximations is also provided therein. Further information is
available upon request.

[1] CMS Collaboration, "Simplified likelihood for the
re-interpretation of public CMS results", CMS Note
CMS-NOTE-2017-001. CERN-CMS-NOTE-2017-001, 2017,
https://cds.cern.ch/record/2242860

1) aggregated_results_crfit.csv

The first line of each .csv file defines the contents of the columns:
njet,btag,htlow,hthigh,mhtlow,mhthigh,bkg,bkgErr,data

Notes: 
- the first 6 columns define the bin in terms of njet, nb, HT, and HTmiss
- bkgdErr contains both stat. and syst. contributions.
- the njet and nb categories are identified through 4-character codes:
  - le2j: njet (symm) == 1, njet (symm) == 2, and njet (asym) == 2
  - ge3a: njet (asym) >= 3  
  - ge3j: njet (symm) == 3 and njet (symm) == 4
  - ge5j: njet (symm) >= 5
  - eq0b: exactly 0 b-tagged jets
  - ge1b: at least 1 b-tagged jet
  - le1b: exactly 0 or 1 b-tagged jets
  - ge2b: at least 2 b-tagged jets
- "symm": implies symmetric pT thresholds of >100GeV on the two leading jets
- "asym": implies asymmetric pT thresholds of >100 and >40GeV on the two leading jets

2) aggregated_results_crfit.root

This file provides identical information to 1), but in a histogram
format (TH1F). The relevant histograms are found in the "shapes_fit_b"
directory and are called "total_data" and "total_background".

In the same file and directory, the covariance matrix is also provided
in a TH2D histogram format ("total_covar"). The matrix can be used in
conjunction with the "result_data" and "total_background" histograms
to form a "simplified likelihood", as detailed in Ref [1] (above) in
order to aid the reinterpretation of the analysis result.

==================================
Code implementations of variables:
==================================

1) makeAlphaT.tar
2) makeBiasedDPhi.tar

1) and 2) contain python snippets that define the two variables used
to suppress QCD multijet production to a negligible level with respect
to other SM backgrounds.