Merge pull request #45774 from AdrianoDee/patch-14

Remove Duplicate RelVals from `relval_highstats.py` and Smaller Queries for `das-up-to-nevents.py`

Configuration/PyReleaseValidation/python/MatrixUtil.py

@@ -133,9 +133,9 @@ def das(self, das_options, dataset):
             command = "dasgoclient %s --query '%s'" % (das_options, self.queries(dataset)[0])
         elif self.skimEvents:
             from os import getenv
-            if getenv("CMSSW_USE_IBEOS","false")=="true": 
+            if getenv("JENKINS_PREFIX") is not None:
                 # to be assured that whatever happens the files are only those at CERN
-                command = "das-up-to-nevents.py -d %s -e %d -s T2_CH_CERN"%(dataset,self.events) 
+                command = "das-up-to-nevents.py -d %s -e %d -pc"%(dataset,self.events)
             else:
                 command = "das-up-to-nevents.py -d %s -e %d"%(dataset,self.events)
         # Run filter on DAS output 

Configuration/PyReleaseValidation/python/relval_data_highstats.py

@@ -16,7 +16,7 @@
 for e_n,era in enumerate(eras_2024):
     for p_n,pd in enumerate(pds_2024):
         for e_key,evs in event_steps_dict.items():
-            if "50k" == e_key: # already defined in relval_standard
+            if "10k" == e_key: # already defined in relval_standard
                 continue   
             wf_number = base_wf_number_2024
             wf_number = wf_number + offset_era * e_n

Configuration/PyReleaseValidation/python/relval_highstats.py

@@ -86,29 +86,3 @@
 workflows[134.99601] = ['',['RunJetHT2015HLHS','HLTDR2_25ns','RECODR2_25nsreHLT_HIPM','HARVESTDR2']]
 workflows[134.99602] = ['',['RunZeroBias2015HLHS','HLTDR2_25ns','RECODR2_25nsreHLT_HIPM','HARVESTDR2']]
 workflows[134.99603] = ['',['RunSingleMu2015HLHS','HLTDR2_25ns','RECODR2_25nsreHLT_HIPM','HARVESTDR2']]
-
-
-
-## 2024 Data Higher Stats Workflows
-## with 150k, 250k, 500k or 1M events each
-
-base_wf_number_2024 = 2024.0
-offset_era = 0.1 # less than 10 eras
-offset_pd = 0.001 # less than 100 pds
-offset_events = 0.0001 # less than 10 event setups (50k,150k,250k,500k)
-
-for e_n,era in enumerate(eras_2024):
-    for p_n,pd in enumerate(pds_2024):
-        for e_key,evs in event_steps_dict.items():
-            if "50k" in e_key: # already defined in relval_standard
-                continue   
-            wf_number = base_wf_number_2024
-            wf_number = wf_number + offset_era * e_n
-            wf_number = wf_number + offset_pd * p_n
-            wf_number = wf_number + offset_events * evs 
-            wf_number = round(wf_number,6)
-            step_name = "Run" + pd + era.split("Run")[1] + "_" + e_key
-            workflows[wf_number] = ['',[step_name,'HLTDR3_2024','AODNANORUN3_reHLT_2024','HARVESTRUN3_2024']]
-
-
-

Configuration/PyReleaseValidation/python/relval_standard.py

@@ -559,19 +559,21 @@
 workflows[142.901] = ['',['RunUPC2023','RECODR3_2024_UPC','HARVESTDPROMPTR3']]
 workflows[142.902] = ['',['RunUPC2023','RECODR3_2024_HIN','HARVESTDPROMPTR3']]
 
-## 2024 Data Workflows
+## 2024 Data Workflows 
+# for a limited set of eras and PDs not to overflow the IB matrices
+#
 base_wf_number_2024 = 2024.0
 offset_era = 0.1 # less than 10 eras
 offset_pd = 0.001 # less than 100 pds
 
-for e_n,era in enumerate(eras_2024):
-    for p_n,pd in enumerate(pds_2024):
+for e_n,era in enumerate(['Run2024D','Run2024C']):
+    for p_n,pd in enumerate(['JetMET0','ZeroBias']):
         wf_number = base_wf_number_2024
         wf_number = wf_number + offset_era * e_n
         wf_number = wf_number + offset_pd * p_n
-        wf_number = wf_number + 0.0001 * 0.05 
+        wf_number = wf_number + 0.0001 * 0.01 
         wf_number = round(wf_number,6)
-        step_name = "Run" + pd + era.split("Run")[1] + "_50k"
+        step_name = "Run" + pd + era.split("Run")[1] + "_10k"
         workflows[wf_number] = ['',[step_name,'HLTDR3_2024','AODNANORUN3_reHLT_2024','HARVESTRUN3_2024']]
 
 ### fastsim ###

Configuration/PyReleaseValidation/python/relval_steps.py

@@ -45,8 +45,8 @@
 steps = Steps()
 
 #### Event to runs
-event_steps = [0.05,0.15,0.25,0.5,1] #in millions
-event_steps_k = ["50k","150k","250k","500k","1M"]
+event_steps = [0.01,0.05,0.15,0.25,0.5,1] #in millions
+event_steps_k = ["10k","50k","150k","250k","500k","1M"]
 event_steps_dict = dict(zip(event_steps_k,event_steps))
 #### Production test section ####
 steps['ProdMinBias']=merge([{'cfg':'MinBias_8TeV_pythia8_TuneCUETP8M1_cff','--relval':'9000,300'},step1Defaults])

Configuration/PyReleaseValidation/scripts/das-up-to-nevents.py

@@ -40,10 +40,9 @@ def das_file_site(dataset, site):
 
 def das_file_data(dataset,opt=""):
     cmd = "dasgoclient --query='file dataset=%s %s| grep file.name, file.nevents'"%(dataset,opt)
-
     out = das_do_command(cmd)
     out = [np.array(r.split(" "))[[0,3]] for r in out if len(r) > 0]
-    
+
     df = pd.DataFrame(out,columns=["file","events"])
     df.events = df.events.values.astype(int)
 
@@ -59,6 +58,28 @@ def das_lumi_data(dataset,opt=""):
 
     return df
 
+def das_run_events_data(dataset,run,opt=""):
+    cmd = "dasgoclient --query='file dataset=%s run=%s %s | sum(file.nevents) '"%(dataset,run,opt)
+    out = das_do_command(cmd)[0]
+
+    out = [o for o in out.split(" ") if "sum" not in o]
+    out = int([r.split(" ") for r in out if len(r)>0][0][0])
+
+    return out
+
+def das_run_data(dataset,opt=""):
+    cmd = "dasgoclient --query='run dataset=%s %s '"%(dataset,opt)
+    out = das_do_command(cmd)
+
+    return out
+
+def no_intersection():
+    print("No intersection between:")
+    print(" - json   : ", best_json)
+    print(" - dataset: ", dataset)
+    print("Exiting.")
+    sys.exit(1)
+
 if __name__ == '__main__':
 
     parser = argparse.ArgumentParser()
@@ -69,6 +90,7 @@ def das_lumi_data(dataset,opt=""):
     parser.add_argument('--pandas', '-pd',action='store_true',help="Store the whole dataset (no event or threshold cut) in a csv") 
     parser.add_argument('--proxy','-p', help='Allow to parse a x509 proxy if needed', type=str, default=None)
     parser.add_argument('--site','-s', help='Only data at specific site', type=str, default=None)
+    parser.add_argument('--precheck','-pc', action='store_true', help='Check run per run before building the dataframes, to avoid huge caching.')
     args = parser.parse_args()
 
     if args.proxy is not None:
@@ -77,6 +99,8 @@ def das_lumi_data(dataset,opt=""):
         print("No X509 proxy set. Exiting.")
         sys.exit(1)
 
+    ## Check if we are in the cms-bot "environment"
+    testing = "JENKINS_PREFIX" in os.environ
     dataset   = args.dataset
     events    = args.events
     threshold = args.threshold
@@ -97,6 +121,7 @@ def das_lumi_data(dataset,opt=""):
     cert_path = base_cert_path + cert_type + "/"
     web_fallback = False
 
+    ## if we have access to eos we get from there ...
     if os.path.isdir(cert_path):
         json_list = os.listdir(cert_path)
         if len(json_list) == 0:
@@ -105,7 +130,7 @@ def das_lumi_data(dataset,opt=""):
         json_list = [c for c in json_list if c.startswith("Cert_C") and c.endswith("json")]
     else:
         web_fallback = True
-
+    ## ... if not we go to the website
     if web_fallback:
         cert_url = base_cert_url + cert_type + "/"
         json_list = get_url_clean(cert_url).split("\n")
@@ -132,12 +157,39 @@ def das_lumi_data(dataset,opt=""):
             R = R + [f for f in range(r[0],r[1]+1)]
         golden_flat[k] = R
 
+    # let's just check there's an intersection between the
+    # dataset and the json
+    data_runs = das_run_data(dataset)
+    golden_data_runs = [r for r in data_runs if r in golden_flat]
+
+    if (len(golden_data_runs)==0):
+        no_intersection()
+
     # building the dataframe, cleaning for bad lumis
-    df = das_lumi_data(dataset).merge(das_file_data(dataset),on="file",how="inner") # merge file informations with run and lumis
-    df = df[df["run"].isin(list(golden.keys()))] # skim for golden runs
+    golden_data_runs_tocheck = golden_data_runs
+    das_opt = ""
+    if testing or args.precheck:
+        golden_data_runs_tocheck = []
+        # Here we check run per run.
+        # This implies more dasgoclient queries, but smaller outputs
+        # useful when running the IB/PR tests not to have huge
+        # query results that have to be cached.
+
+        sum_events = 0
+
+        for r in golden_data_runs:
+            sum_events = sum_events + int(das_run_events_data(dataset,r))
+            golden_data_runs_tocheck.append(r)
+            if events > 0 and sum_events > events:
+                break
+
+        das_opt = "run in %s"%(str([int(g) for g in golden_data_runs_tocheck]))
+
+    df = das_lumi_data(dataset,opt=das_opt).merge(das_file_data(dataset,opt=das_opt),on="file",how="inner") # merge file informations with run and lumis
+
     df["lumis"] = [[int(ff) for ff in f.replace("[","").replace("]","").split(",")] for f in df.lumis.values]
     df_rs = []
-    for r in golden_flat:
+    for r in golden_data_runs_tocheck:
         cut = (df["run"] == r)
         if not any(cut):
             continue
@@ -152,12 +204,8 @@ def das_lumi_data(dataset,opt=""):
         n_lumis = np.array([len(l) for l in df_r.lumis])
         df_rs.append(df_r[good_lumis==n_lumis])
 
-    if len(df_rs) == 0:
-        print("No intersection between:")
-        print(" - json   : ", best_json)
-        print(" - dataset: ", dataset)
-        print("Exiting.")
-        sys.exit(1)
+    if (len(df_rs)==0):
+        no_intersection()
 
     df = pd.concat(df_rs)
     df.loc[:,"min_lumi"] = [min(f) for f in df.lumis]