Merge pull request cms-sw#75 from CMS-HGCAL/hgcal-condformat-HGCalNAN…

…O-13_2_0_pre3_feat

add example

DPGAnalysis/HGCalTools/test/MIPPeakSimple.py

@@ -0,0 +1,116 @@
+import os
+import ROOT
+import pandas as pd
+import numpy as np
+import re
+
+basedir='/eos/cms/store/group/dpg_hgcal/comm_hgcal/2023/CMSSW/ReReco_Sep26/'
+pedestal_run=f'{basedir}/Run1695716961/d1643b06-5cc3-11ee-bee0-fa163e17e03d'
+#mip_run=f'{basedir}/Run1695563673/c6adb2e6-5cc3-11ee-bee0-fa163e17e03d'
+#mip_run=f'{basedir}/Run1695564190/c713c360-5cc3-11ee-bee0-fa163e17e03d'
+mip_run=f'{basedir}/Run1695564694/c764c29c-5cc3-11ee-bee0-fa163e17e03d'
+#mip_run=f'{basedir}/Run1695565177/c7d12c48-5cc3-11ee-bee0-fa163e17e03d'
+#mip_run=f'{basedir}Run1695565613/c82c5db6-5cc3-11ee-bee0-fa163e17e03d'
+selection_metric='NoiseRatio'
+
+def findCandidateChannels(ped,mip,metric='NoiseRatio'):
+
+    """ reads the DQM outputs and selects the channel for which std(mip)/std(ped) has increased the most """
+
+
+    def _setChannelModule(df):
+
+        """helper function to uniformize channel"""
+
+        #string to integer
+        df['Channel'] = df['Channel'].apply(lambda x: int(x, 16))
+
+        #in case it's not a DTH run the modules come as two separate FEDs
+        df['Channel'] = np.where(df['Channel']>0x40000,
+                                     df['Channel']-0x40000+0x400,
+                                     df['Channel'])
+
+        #assign a module number fo grouping
+        df['Module'] = np.where((df['Channel']<0x400),1,2)
+        df=df.astype({'Module':int})
+
+        return df
+
+
+    #merge mip and pedestal summaries
+    columns=['Channel','Pedestal','Noise']
+    ped_df=pd.read_csv(ped,sep='\s+',header='infer')[columns]
+    ped_df=_setChannelModule(ped_df)
+    mip_df=pd.read_csv(mip,sep='\s+',header='infer')[columns]
+    mip_df=_setChannelModule(mip_df)
+    df=mip_df.merge(ped_df,on='Channel',suffixes=('', '_ped'))
+
+    #remove dead channels
+    mask=(df['Noise']==0) | (df['Noise_ped']==0)
+    df=df[~mask].copy()
+    df['NoiseRatio'] = df['Noise']/df['Noise_ped']
+
+    #select channels of interest
+    max_var_channels = df.groupby(['Module'])[metric].idxmax().values
+    mip_spot_channels = df.iloc[max_var_channels][['Channel','Module','Pedestal_ped','NoiseRatio','Noise']]
+    print(f'Identified two possible candidates for MIP spots by {metric}')
+    print(mip_spot_channels.head())
+
+    return mip_spot_channels
+
+
+mip_spot_channels=findCandidateChannels(ped=f'{pedestal_run}/calibs/level0_calib_params.txt',
+                                        mip=f'{mip_run}/calibs/level0_calib_params.txt',
+                                        metric=selection_metric)
+chlist=mip_spot_channels['Channel'].values
+pedlist=mip_spot_channels['Pedestal_ped'].values
+modlist=mip_spot_channels['Module'].values
+
+#select in NANOAOD, subtract pedestals and draw the MIP peak
+flist=[os.path.join(mip_run,f) for f in os.listdir(mip_run) if 'NANO' in f]
+
+ROOT.ROOT.EnableImplicitMT()
+
+df=ROOT.RDataFrame('Events',flist)
+
+chlist=mip_spot_channels['Channel'].values
+chsel = ' || '. join( [f'HGC_eleid=={ch}' for ch in chlist] )
+df=df.Define('goodch',chsel).Filter('Sum(goodch)==2')
+
+pedlist=mip_spot_channels['Pedestal_ped'].values
+histos=[]
+for i in range(len(chlist)):
+    pedval=pedlist[i]
+    h = df.Define(f'goodch_ADC{i}',f'HGC_adc[goodch][{i}]-{pedval}') \
+          .Histo1D( (f'adc{i}',f'Module {modlist[i]};ADC counts;Events',50,0,50), f'goodch_ADC{i}')
+    histos.append(h.GetPtr())
+
+ROOT.ROOT.DisableImplicitMT()
+
+
+#plot
+ROOT.gStyle.SetOptStat(0); 
+ROOT.gStyle.SetTextFont(42)
+ROOT.gStyle.SetOptTitle(0)
+c = ROOT.TCanvas("c", "c", 800, 800)
+c.SetTopMargin(0.05)
+c.SetLeftMargin(0.12)
+c.SetRightMargin(0.05)
+frame=histos[0].Clone('frame')
+frame.Reset('ICE')
+frame.Draw()
+run_number=re.findall('Run(\d+)',mip_run)[0]
+leg=ROOT.TLegend(0.6,0.7,0.95,0.95,f'Run {run_number}')
+leg.SetFillStyle(0)
+leg.SetBorderSize(0)
+for i,h in enumerate(histos):
+    h.Draw('histsame')
+    h.SetLineWidth(2)
+    h.SetLineColor(i+1)
+    leg.AddEntry(h,h.GetTitle(),'l')
+frame.GetYaxis().SetRangeUser(1,1e6)
+c.SetLogy()
+leg.Draw()
+c.Modified()
+c.Update()
+c.SaveAs(f'/eos/user/p/psilva/www/HGCAL/NanoAOD/mip_{run_number}_{selection_metric}.png')