TODOs

Artus

HP

• make the creation of these plots possible
  • the default root plot module only supports one subplot canvas and Thomas is not sure, whether we should add more canvases here (because this duplicates many settings). He would propose to start with a new plotting modules to be run after the current one
• fix the automatic width adjustion for the option --function-display-result in FunctionPlot module. Example:

higgsplot.py -f "demo/once_tree" \
-x v_v0_Ks_inv_m_pi \
--live -i out_mc_full_finer_binning_new.root \
--nicks mc_nick -m LINE \
--x-label "Inv. mass of K daughters, GeV" --y-label "norm. entries" \
--analysis-modules NormalizeToUnity FunctionPlot \
--functions "[0]*(exp(-0.5*((x-[1])/[2])**2))  + [3]*x+[4]" \
--function-parameters   0.06,0.49769,0.003475,0.017,0.0051458 \
--function-ranges 0.45,0.55 --x-lims 0.4 0.6 \
--function-fit mc_nick \
--log-level debug \
--x-bins 1000,0.4,0.6 \
--function-display-result \
--function-fit-parameter-names "a" "b" "c" "d" "e"

• Find an alternative or a fix for web_plotting_no_passwd such that it also works for using webplotting for EOS-sites (see also this commit made on the corresponding branch).
• Fix the ignoring of bin susing log-scale for 3D histograms. Example:

higgsplot.py -f "demo/kaon_tree" \
-z v0_Ks_inv_m_pi -x hps_pt_1 -y hps_pt_2 \
-w "v0_Ks_inv_m_pi>0 && hps_pt_1>0" \
--live -i out_mc_full_matched_byReference.root \
--nicks mc_v0_Ks1 \
--z-bins 400,0.4,0.6 --x-bins "0 3 8 1000" --y-bins "0 3 8 1000" --z-lims 0.4 0.6 --x-lims 0.1 10 --y-lims 0.1 10  --x-log  --y-log \
-m TEXTCOLZ \
--filename projection \
--analysis-modules NormalizeToUnity ProjectByFit \
--projection-functions "[0]*(exp(-0.5*((x-[1])/[2])**2))  + [3]*x+[4]" \
--projection-parameters 0.00615232,0.497674,0.00434818,0.00172523,0.00009 \
--projection-to-plot 1 \
--projection-to-nick mc_v0_Ks1 \
--projection-fit-range 0.4,0.6 \
--projection-fit-slices-z

Htt

• Remove duplicated quantities from Artus output ntuples
  • Prefer LV quantities over float components
• Gen/RecoTauCPProducer: some refactoring is needed (e.g. if statements, naming convention, remove not needed functions)

Kappa

• can we remove datasetHelper? Only datasetHelperTwopz is used. I also don’t see it in Artus and Htt where in the later datasetsHelper2015 is used (only isData). datasetsHelper2015 can be probably removed too.
  • fix the oppended issue on dublicating functionality
• properly use getValidVertexIndex in KVertexSummary producer

KappaTools

• remove not-used blocks from the checkout scripts

Tests

• Add the Artus test to travis. Requires discussion on how we want it. Seems like Raphael already is doing some steps towards this

Additional tools

• compareRootFiles.py in Htt with option -a creates error message if the number of leafs is different. It should be possible to compare the branches that match