Writing MC true instead of MC observed

Slightly modified plot illustrating the composition of our background prediction (Bpred_Data_____PredictionComposition): now jzb<0 is below sidebands/emu

Updated 'Bpred_Data_____PredictionComposition' (i.e. composition of the prediction) to be more comprehensible and less busy

Adapted range for efficiency_curve_jzb and new option in DrawPrelim. If the input argument is 0 no luminosity tag is added to the text.

Also saving JZB data vs mc for JZB>0 only (instead of full range) to compare it to Bpred plot; using mcjzb variable for mc, datajzb variable for data

Several new features: Bpred_Data now has a new version showing the two contributions: Z+Jets and TTbar prediction; several plots now have fixed binning size that makes sense (i.e. 5 GeV bins instead of like 5.2398754839 GeV); enabled SMS signal vs bg plot (loading all SMS zones, and unloading them) [keep in mind that you need to have ScanSampleDirectory set to 'SMS/' in Setup.C for this to work in the first place]

Fixed lines for mll plot (showing the Z peak window); added a plot showing the composition of the background prediction (for data) in a colorful way; adapted lower boundary of Bpred plots (MC and MC w/ S)

Moved peak finding without PURW for z+jets systematic into the corresponding function to tidy up a bit

Forcing the x axis range for JZB plot (with ratio); adapted the line position for the ttbar systematic uncertainty plot; reset mll comparison plot x range

Adapted binning for ttbar systematics plot so tails become more easily interpretable

Always producing ratio plot for kin plots

Added JZB efficiency plot and possibility to get peak without PU reweighting

Reduced verbose compiling output (e.g. double instead of float and such)

Saving to template updated for BZJ

Ratio plot (with jzb ratio binning) directly on bpred plot; adapted draw_ratio_on_canvas to more intuitive save_with_ratio (which saves the canvas along with a canvas also containing the ratio [suffix _withratio); implemented saving of flipped template; and some extras ...

Writing 'Opposite flavor' instead of 'OSOF' and 'Same flavor' instead of 'OSSF'; also adapted legend size for those two cases so the font is the same size

Also draw background only background prediction on b+s background prediction plot

Can now easily switch between versions

Potential memory leak fix conditional: only needed when RTMP

Added option which automatically generates the table with the total number of events observed in the auxiliary region |JZB|>30 GeV u |JZB|< 50 GeV for the OFZP and OFSB

Adapted ratio plot for ttbar sideband comparison to make it more easily readable

Drawing lines on mll plots to illustrate sidebands

Updated todo list

Reverted back to fixed binning, but use 50GeV bins.

Passing ratio binning to sideband comparison

Also doing ratio plot for m_{ee} and m_{mm}; for response plot in MC explicitly writing CMS MC Preliminary (...)

Saving the y scale for the mll (ossf) plot, and reusing it for the osof mll plot

Updated y range for OSOF mll plots

Added a variable (RestrictToMassPeak) to switch off peak analysis option on and off; at the moment there are only warnings as placeholders in order to avoid any possible interference with the current (pre)approval workflow

Adapted ttbar sideband comparison legend; providing good start parameters for Bpred fit function that can be used instead of partial fitting to go right to the final fitting

For plots without PURW the weight is now automatically set to 1 (no need to change the config every time); the n(vtx) plot now goes up to 20; ratio legend is smaller;

simulated lumi defaults to official lumi

Now also drawing LM4 on comparison plots

Use 1/fb as benchmark for signal vs. bkgd comparison.

Adapted y axis of comparison plots for 2 /fb

Removing doublechecks and drawing lm4

Changed the search name for data in the response correction as this could have triggered a problem when replacing the data sample (possible segfault)

Set minimum in Bpred so that 1 event in one of the control regions still appears.
Also, always plot observed distribution on top (re-drawing).

Adapted signal vs background plots so they are normalized to 1 /fb

Adapted ttbar comparison plot in accordance with the official notation

Added ttbar comparison plot (sidebands vs peak etc.)

Adding several features and cleaning up a bit: \n - Added a new analytical description of the prediction shape, based on a Z+Jets crystal ball fit and a ttbar logpar fit; the two are combined (see code for more info or ask me) \n - Changed name of template function (which essentially writes a template of the results to a file for use in the context of simplified model scans \n - Updated todo list (will hopefully get less entries soon) \n - Removed some unncessary lines

First commits towards Poisson ratio plots

Fixed two potential memory leaks

Adapted response correction function; basically it now calculates the response in one go directly from the fit to the histo that gets drawn; furthermore, the result gets written on the canvas

Now providing the option to run ./Create_All_Plots.exec 1 to calculate peak positions, observed and predicted yields, and save all information in a file that can then be used for the susy model scans; this means that all these steps are just once and the data can then be used, instead of having each job redo this.

Getting rid of some potential memory leaks (due to cloning) and a seg fault

Improved response plot.
Improved region comparison plots.

Several updates for the AN:
- Implement name and color conventions for the control and signal regions;
- improved legends and generally plots' style;
- add switch to overlay LM4 signal or not;
- make sure plots have the same scale (e.g., by always plotting data first);
- removed 1-sigma poisson envelope in Bpred plot (needs some work);
- harmonized x-axis ranges;
- added function for ZP region [is_ZP(TCut& cut)];

Added a todo function to keep sight of tasks ahead

Moved different limit calculation functions to separate module to unclutter Plotting_Functions; moved unneeded functions to the dusty module

Added more plots (mll split up for ee/mm); fixed one potential memory leak

Updated style: exponent is now legible, cut isn't drawn on top anymore. CMS Preliminary 2011 (etc.) is now right above the plot, allowing the legend to take up more space.

Response plot now with log z

Added several plots to understand sidebands better; made sideband definition global (now in Setup.C)

Updated the ttbar extraction function to get rid of old stuff and take advantage of the new cb_fit function

Changed ttbar test algorithm as well; Introduced correction term for jzb (based on pt)

TTbar presentation update

Taking signal contamination into account

Looking into Z+Jets and TTbar shapes (from data); Added band for LogParabola