From 184bee16cd70e94607f93b0644b3c567ada54b51 Mon Sep 17 00:00:00 2001
From: Laurent Mercadier <laurent.mercadier@xfel.eu>
Date: Mon, 19 Jun 2023 12:13:49 +0200
Subject: [PATCH] Adds function to check data rate (missing trains)
---
src/toolbox_scs/load.py | 47 +++++++++++++++++++++++++++++++++++++++++
1 file changed, 47 insertions(+)
diff --git a/src/toolbox_scs/load.py b/src/toolbox_scs/load.py
index 4edbcc3..a28f1c4 100644
--- a/src/toolbox_scs/load.py
+++ b/src/toolbox_scs/load.py
@@ -31,6 +31,7 @@ __all__ = [
'open_run',
'run_by_path',
'load_run_values',
+ 'check_data_rate',
]
log = logging.getLogger(__name__)
@@ -209,6 +210,7 @@ def load(proposalNB=None, runNB=None,
'Skipping!')
continue
data_arrays.append(arr)
+ # Check missing trains
for arr in data_arrays:
rate = arr.sizes["trainId"] / len(run.train_ids)
if rate < 0.95:
@@ -494,3 +496,48 @@ def concatenateRuns(runs):
for k in orderedRuns[0].attrs.keys():
result.attrs[k] = [run.attrs[k] for run in orderedRuns]
return result
+
+
+def check_data_rate(run, mnemonics=None):
+ """
+ Calculates the fraction of train ids that contain data in a run.
+
+ Parameters
+ ----------
+ run: extra_data DataCollection
+ the DataCollection associated to the data.
+ mnemonics: str, list of str or dict
+ mnemonics to check. If None, all mnemonics in the run are checked.
+ A custom mnemonic can be defined with a dictionnary: {'extra':
+ {'source': 'SCS_CDIFFT_MAG/SUPPLY/CURRENT', 'key':
+ 'actual_current.value'}}
+ Output
+ ------
+ ret: dictionnary
+ dictionnary with mnemonic as keys and fraction of train ids
+ that contain data as values.
+ """
+ run_mnemonics = mnemonics_for_run(run)
+ if mnemonics is None:
+ mnemonics = run_mnemonics
+ mnemonics = [mnemonics] if isinstance(mnemonics, str) else mnemonics
+ ret = {}
+ for m in mnemonics:
+ if isinstance(m, dict):
+ name = list(m.keys())[0]
+ val = m[name]
+ m = name
+ elif m not in run_mnemonics:
+ log.warning(f'mnemonic {m} not found. Skipping!')
+ continue
+ else:
+ val = run_mnemonics[m]
+ counts = run[val['source']][val['key']].data_counts(False)
+ npulses = counts.max()
+ if npulses == 0: # (only missing data)
+ rate = 0.
+ else:
+ counts = counts / npulses # to only count trains and not pulses
+ rate = counts.sum() / len(run.train_ids)
+ ret[m] = rate
+ return ret
--
GitLab