def get_data(q):
api = RhApi(DEFAULT_URL, debug = False)
p = {"class": "Collisions18"}
qid = api.qid(q)
return api.csv(q, p)
def retrieve(max_run=320008,
min_run=316766,
folder="runreg_csc",
table="datasets",
ref_runs=[]):
api = RhApi(DEFAULT_URL, debug=False)
# Get column names and name of run column
col_table = api.table(folder=folder, table=table)["columns"]
cols = []
r_num = ""
r_num_i = 0
for col in col_table:
col_name = str(col["name"])
if _get_data_col(col_name, table): cols.append(col_name)
if not r_num and _get_run_col(col_name):
cols.append(col_name)
r_num = col_name
r_num_i = cols.index(col_name)
if not r_num:
r_num = cols[0]
# Form query
c = ",".join("r." + x for x in cols)
q = "select {0} from {1}.{2} r where r.{3}>=:minrun and r.{3}<=:maxrun order by r.{3}".format(
c, folder, table, r_num)
if ref_runs:
p = {"maxrun": max(ref_runs), "minrun": min(ref_runs)}
else:
p = {"maxrun": str(max_run), "minrun": str(min_run)}
qid = api.qid(q)
# Allow for fetch() to update a pre-existing dict
data = {}
if type(ref_runs) == dict:
dqm = ref_runs
ref_runs = ref_runs.keys()
skipped = 0
it = 0
while True:
runs = []
raw_data = api.json(q, p)["data"]
for i in range(0, len(raw_data)):
run = str(raw_data[i][r_num_i])
# Only fetch runs relevant to AutoDQM
if run not in ref_runs: continue
# Get source of data (Global, Express, or PromptReco)
if "RDA_NAME" in cols:
rda_name = raw_data[i][cols.index("RDA_NAME")].lower()
runs.append(run)
# Only make a new entry for new runs
if run not in data:
data[run] = {}
for j in range(0, len(raw_data[i])):
if j == r_num_i: continue
# Handle <folder>.datasets
if table != "runs":
if "is_good" not in data[run]: pass
elif not data[run]["is_good"]: continue
else:
# The "BAD" tag gets priority
if raw_data[i][j] == "BAD":
data[run]["is_good"] = False
# Skip "NONSET" tags if marked as "GOOD" elsewhere
continue
if "RDA_NAME" in data[run]:
# Update source
data[run]["RDA_NAME"] = rda_name
# Update status
if raw_data[i][j] == "GOOD": data[run]["is_good"] = True
elif raw_data[i][j] == "BAD": data[run]["is_good"] = False
else: continue
# Handle <folder>.runs
else:
data[run][cols[j]] = raw_data[i][j]
if table != "runs":
# handle cases where all statuses are NOTSET
if "is_good" not in data[run]: data[run]["is_good"] = False
if len(raw_data) < 1 or max(
runs) == p["minrun"] or p["minrun"] >= p["maxrun"]:
break
p["minrun"] = max(runs)
it += 1
if table == "runs" and ref_runs:
refs = {"ref_data": [], "ref_cands": []}
for run in data:
if run == max(ref_runs): continue
refs["ref_data"].append(
dict(ref.get_wbm_data(max(ref_runs), run, data), **dqm[run]))
refs["ref_cands"] = ref.get_ref_cands(refs["ref_data"])
return refs
elif data:
return data
else:
return None