Update for D+ jets (alisw#1009)

* Use latest Dplus trains

* Update DB for Dplus

* First version for Dplus

* Adjust ptjet bins for Dplus

* Update DB for D+

* Revert results directory

* Fix formatting

Author: qgp

machine_learning_hep/data/data_run3/database_ml_parameters_DpJet_pp.yml

@@ -278,7 +278,7 @@ DpJet_pp:
       seedmerge: [12] #list of periods
       period: [LHC23] #list of periods
       select_period: [1]
-      prefix_dir: /data2/MLhep/real/train_347181/
+      prefix_dir: /data2/MLhep/trains/364134/
       unmerged_tree_dir: [alice] #list of periods
       pkl: ["${USER}/dpjet/pkl"] #list of periods
       pkl_skimmed: ["${USER}/dpjet/pklsk"] #list of periods
@@ -295,7 +295,7 @@ DpJet_pp:
       seedmerge: [12] #list of periods
       period: [LHC24h1] #list of periods
       select_period: [1]
-      prefix_dir: /data2/MLhep/sim/train_348146/
+      prefix_dir: /data2/MLhep/trains/364345/
       unmerged_tree_dir: [alice]
       pkl: ["${USER}/dpjet/pkl"] #list of periods
       pkl_skimmed: ["${USER}/dpjet/pklsk"] #list of periods
@@ -377,8 +377,8 @@ DpJet_pp:
     jet_obs: &jet_default
       sel_an_binmin: [2, 3, 4, 5, 6, 7, 8, 10, 12, 16] # hadron pt bins (sel_an_binmin bins)
       sel_an_binmax: [3, 4, 5, 6, 7, 8, 10, 12, 16, 24] # hadron pt bins (sel_an_binmin bins)
-      bins_ptjet: [2, 5, 7, 10, 15, 30] # systematics, TODO: split rec and gen binning
-      bins_ptjet_eff: [2, 5, 7, 15, 30, 50] # systematics, TODO: split rec and gen binning
+      bins_ptjet: [5, 7, 10, 15, 20, 30, 40, 60, 70]
+      bins_ptjet_eff: [2, 5, 7, 10, 15, 20, 30, 40, 60, 70, 90]
       cand_collidx: fIndexHFDPCOLLBASES
       counter_read_data: fReadCountsWithTVXAndZVertexAndSel8
       counter_read_mc: fReadCountsWithTVXAndZVertexAndSelMC
@@ -389,45 +389,45 @@ DpJet_pp:
 
       observables:
         zg:
-          bins_gen_fix: [6, -.1, .5]
-          bins_det_fix: [6, -.1, .5]
+          bins_gen_var: [-.1, .0, .05, .1, .15, .2, .25, .3, .35, .4, .45, .5]
+          bins_det_var: [-.1, .0, .05, .1, .15, .2, .25, .3, .35, .4, .45, .5]
           label: "#it{z}_{g}"
           label_y: "(1/#it{N}_{jet ch}) d#it{N}/d#it{z}_{g}"
         nsd:
-          bins_gen_fix: [10, -.5, 9.5]
-          bins_det_fix: [10, -.5, 9.5]
+          bins_gen_fix: [7, -.5, 6.5]
+          bins_det_fix: [7, -.5, 6.5]
           label: "#it{n}_{SD}"
           label_y: "(1/#it{N}_{jet ch}) d#it{N}/d#it{n}_{SD}"
         rg:
-          bins_gen_fix: [11, -.1, 1.]
-          bins_det_fix: [11, -.1, 1.]
+          bins_gen_var: [-.1, .0, .05, .1, .15, .2, .25, .3, .35, .4, .45, .5]
+          bins_det_var: [-.1, .0, .05, .1, .15, .2, .25, .3, .35, .4, .45, .5]
           label: "#it{R}_{g}"
           label_y: "(1/#it{N}_{jet ch}) d#it{N}/d#it{R}_{g}"
         zpar:
           # bins_gen_fix: [10, 0., 1.]
           # bins_det_fix: [10, 0., 1.]
-          bins_gen_var: [0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.]
-          bins_det_var: [0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.]
+          bins_gen_var: [0.2, 0.25, 0.3, 0.35, 0.4, 0.45, 0.5, 0.55, 0.6, 0.65, 0.7, 0.75, 0.8, 0.85, 0.9, 0.95, 1.]
+          bins_det_var: [0.2, 0.25, 0.3, 0.35, 0.4, 0.45, 0.5, 0.55, 0.6, 0.65, 0.7, 0.75, 0.8, 0.85, 0.9, 0.95, 1.]
           label: "#it{z}_{#parallel}"
           label_y: "(1/#it{N}_{jet ch}) d#it{N}/d#it{z}_{#parallel}"
         dr:
           bins_gen_fix: [10, 0., 1.]
           bins_det_fix: [10, 0., 1.]
           label: "#Delta#it{r}"
-        lntheta:
-          bins_gen_fix: [10, 0., 5.]
-          bins_det_fix: [10, 0., 5.]
-          label: "#minusln(#it{#theta})"
-          arraycols: [3]
-        lnkt:
-          bins_gen_fix: [10, -8., 2.]
-          bins_det_fix: [10, -8., 2.]
-          label: "ln(#it{k}_{T}/(GeV/#it{c}))"
-          arraycols: [3]
-        lntheta-lnkt:
-          arraycols: [3, 4]
-
-      n_rebin: [3, 3, 3, 3, 3, 3, 3, 4, 5, 5]
+          # lntheta:
+          #   bins_gen_fix: [10, 0., 5.]
+          #   bins_det_fix: [10, 0., 5.]
+          #   label: "#minusln(#it{#theta})"
+          #   arraycols: [3]
+          # lnkt:
+          #   bins_gen_fix: [10, -8., 2.]
+          #   bins_det_fix: [10, -8., 2.]
+          #   label: "ln(#it{k}_{T}/(GeV/#it{c}))"
+          #   arraycols: [3]
+          # lntheta-lnkt:
+          #   arraycols: [3, 4]
+
+      # n_rebin: [3, 3, 3, 3, 3, 3, 3, 4, 5, 5]
       pdf_names:
         pdf_sig: "sig"
         pdf_bkg: "bkg"
@@ -445,103 +445,109 @@ DpJet_pp:
         - level: mc
           range: [1.70, 1.99]
           components:
-            sig:
-              fn: 'Gaussian::sig(m[1., 4.], mean[1.8, 1.92], sigma_g1[.006,.006,.025])'
-            #wide:
-            # fn: 'Gaussian::wide(m, mean, expr("n*sigma_g1", n[1.,5.], sigma_g1))'
-            #model:
-            #  fn: 'SUM::sig(f_peak[0.,1.]*peak, wide)'
-            bkg:
-              fn: 'Exponential::mcbkg(m, mcalpha[0.])'
             model:
-              fn: 'SUM::mctot(mcfrac[0., 0., 1.0]*sig, mcbkg)'
+              fn: 'RooCrystalBall::sig(m[1., 4.], mean[1.87, 1.85, 1.89], sigma_g1[0.006, 0.035], alpha1[0., 2.], n1[0., 100.], alpha1, n1)'
+
+            # sig:
+            #   fn: 'Gaussian::sig(m[1., 4.], mean[1.85, 1.89], sigma_g1[.006,.006,.035])'
+            # wide:
+            #   fn: 'Gaussian::wide(m, mean, expr("n*sigma_g1", n[1.,200.], sigma_g1))'
+            # model:
+            #  fn: 'SUM::sig(f_peak[1., 0.,1.]*peak, wide)'
+
+            # bkg:
+            #   fn: 'Exponential::mcbkg(m, mcalpha[0.])'
+            # model:
+            #   fn: 'SUM::mctot(mcfrac[0., 0., 1.0]*sig, mcbkg)'
         - ptrange: [1., 2.]
           range: [1.70, 1.99]
-          #fix_params: ['n', 'f_peak']
+          # fix_params: ['n', 'f_peak']
+          fix_params: ['alpha1', 'n1']
           components:
             #sig:
             #fn: 'Gaussian::sig(m, mean[2.28,2.29], sigma_g1[.005,.005,.015])'
             bkg:
-              fn: 'Polynomial::bkg(m, {a0[-1e10, 1e10], a1[-1e10, 1e10], a2[-1e10, 1e10]})'
+              fn: 'Polynomial::bkg(m, {a1[0., -3., 3.], a2[0., -3., 3.]})'
             model:
               fn: 'SUM::sum(f_sig[0.,1.]*sig, bkg)'
         - ptrange: [2., 3.]
           range: [1.70, 1.99]
-          #fix_params: ['n', 'f_peak']
+          # fix_params: ['n', 'f_peak']
+          fix_params: ['alpha1', 'n1']
           components:
             # sig:
             #   fn: 'Gaussian::sig(m, mean[2.28,2.29], sigma_g1[.005,.005,.015])'
             bkg:
-              fn: 'Polynomial::bkg(m, {a0[-1e10, 1e10], a1[-1e10, 1e10], a2[-1e10, 1e10]})'
+              fn: 'Polynomial::bkg(m, {a1[0., -10., 10.], a2[10., 0., 10.]})'
             model:
-              fn: 'SUM::sum(f_sig[0.,1.]*sig, bkg)'
+              fn: 'SUM::sum(f_sig[0.,0.,1.]*sig, bkg)'
         - ptrange: [3., 4.]
           range: [1.70, 1.99]
-          #fix_params: ['n', 'f_peak']
+          fix_params: ['n', 'f_peak']
           components:
             #sig:
             #fn: 'Gaussian::sig(m, mean[2.28,2.29], sigma_g1[.005,.005,.015])'
             bkg:
-              fn: 'Polynomial::bkg(m, {a0[-1e10, 1e10], a1[-1e10, 1e10], a2[-1e10, 1e10]})'
+              fn: 'Polynomial::bkg(m, {a1[0., -1., 1.], a2[10., 0., 10.]})'
             model:
               fn: 'SUM::sum(f_sig[0., 0., 0.8]*sig, bkg)'
         - ptrange: [4., 5.]
           range: [1.70, 1.99]
-          #fix_params: ['n', 'f_peak']
+          fix_params: ['n', 'f_peak']
           components:
             # sig:
             #   fn: 'Gaussian::sig(m, mean[2.28,2.29], sigma_g1[.005,.005,.015])'
             bkg:
-              fn: 'Polynomial::bkg(m, {a0[5000, -1e10, 1e10], a1[-1e10, 1e10], a2[-1e10, 1e10]})'
+              fn: 'Polynomial::bkg(m, {a1[0., -100., 100.], a2[10., -10., 10.]})'
             model:
               fn: 'SUM::sum(f_sig[0.,1.]*sig, bkg)'
         - ptrange: [5., 6.]
           range: [1.70, 1.99]
-          #fix_params: ['n', 'f_peak']
+          fix_params: ['n', 'f_peak']
           components:
             # sig:
             #   fn: 'Gaussian::sig(m, mean[2.28,2.29], sigma_g1[.005,.005,.015])'
             bkg:
-              fn: 'Polynomial::bkg(m, {a0[8000, -1e10, 1e10], a1[-1e10, 1e10], a2[-1e10, 1e10]})'
+              fn: 'Polynomial::bkg(m, {a1[0., -100., 100.], a2[10., -10., 10.]})'
             model:
               fn: 'SUM::sum(f_sig[0.,1.]*sig, bkg)'
         - ptrange: [6., 7.]
           range: [1.70, 1.99]
-          #fix_params: ['n', 'f_peak']
+          fix_params: ['n', 'f_peak']
           components:
             # sig:
             #   fn: 'Gaussian::sig(m, mean[2.28,2.29], sigma_g1[.005,.03])'
             bkg:
-              fn: 'Polynomial::bkg(m, {a0[219, -1e10, 1e10], a1[-1e10, 1e10], a2[-1e10, 1e10]})'
+              fn: 'Polynomial::bkg(m, {a1[0., -100., 100.], a2[10., -10., 10.]})'
             model:
               fn: 'SUM::sum(f_sig[0.,1.]*sig, bkg)'
         - ptrange: [7., 8.]
           range: [1.70, 1.99]
-          #fix_params: ['n', 'f_peak']
+          fix_params: ['n', 'f_peak']
           components:
             # sig:
             #   fn: 'Gaussian::sig(m, mean[2.28,2.29], sigma_g1[.005,.03])'
             bkg:
-              fn: 'Polynomial::bkg(m, {a0[200, -1e10, 1e10], a1[-1e10, 1e10], a2[-1e10, 1e10]})'
+              fn: 'Polynomial::bkg(m, {a1[0., -100., 100.], a2[10., -10., 10.]})'
             model:
               fn: 'SUM::sum(f_sig[0.,1.]*sig, bkg)'
         - ptrange: [8., 10.]
           range: [1.70, 1.99]
-          #fix_params: ['n', 'f_peak']
+          fix_params: ['n', 'f_peak']
           components:
             # sig:
             #   fn: 'Gaussian::sig(m, mean[2.28,2.29], sigma_g1[.005,.03])'
             bkg:
-              fn: 'Polynomial::bkg(m, {a0[500, -1e10, 1e10], a1[-1e10, 1e10], a2[-1e10, 1e10]})'
+              fn: 'Polynomial::bkg(m, {a1[0., -100., 100.], a2[10., -10., 10.]})'
             model:
               fn: 'SUM::sum(f_sig[0.,1.]*sig, bkg)'
         - range: [1.70, 2.00]
-          #fix_params: ['n', 'f_peak']
+          fix_params: ['n', 'f_peak']
           components:
             # sig:
             #   fn: 'Gaussian::sig(m, mean[2.28,2.29], sigma_g1[.005,.03])'
             bkg:
-              fn: 'Polynomial::bkg(m, {a0[200, -1e10, 1e10], a1[-1e10, 1e10], a2[-1e10, 1e10]})'
+              fn: 'Polynomial::bkg(m, {a1[0., -100., 100.], a2[10., -10., 10.]})'
             model:
               fn: 'SUM::sum(f_sig[0.,1.]*sig, bkg)'
 
@@ -552,13 +558,13 @@ DpJet_pp:
             signal: [-2., 2.] # systematics
             right: [3., 5.5] # systematics
 
-      mass_fit:
-        func_sig: "gaus"
-        func_bkg: "expo"
-        # par_start:
-        # par_fix: {1: 2.286}
-        par_constrain: {1: [2.28, 2.29], 2: [.005, .03]}
-        range: [1.7, 2.00]
+      # mass_fit:
+      #   func_sig: "gaus"
+      #   func_bkg: "expo"
+      #   # par_start:
+      #   # par_fix: {1: 2.286}
+      #   par_constrain: {1: [2.28, 2.29], 2: [.005, .03]}
+      #   range: [1.7, 2.00]
 
       efficiency:
         index_match: fIndexArrayDPCMCPJETOS_hf

machine_learning_hep/processer_jet.py

@@ -437,6 +437,7 @@ def process_efficiency_single(self, index):
                     "fPairPt",
                 ]
             )
+            cols = None
 
             # read generator level
             dfgen_orig = pd.concat(