From 70351027962d92c93ec7c499451b35d9d538f334 Mon Sep 17 00:00:00 2001
From: Daniel Estrada <daniel.estrada.acevedo@cern.ch>
Date: Thu, 18 May 2023 01:27:25 +0200
Subject: [PATCH] Add S-curve filter to identify noisy channels based on ENC
value
---
gemos/analysis/scurve_analysis.py | 28 ++++++++++++++++++++++++++++
1 file changed, 28 insertions(+)
diff --git a/gemos/analysis/scurve_analysis.py b/gemos/analysis/scurve_analysis.py
index cf7f923..f7bde75 100644
--- a/gemos/analysis/scurve_analysis.py
+++ b/gemos/analysis/scurve_analysis.py
@@ -7,6 +7,7 @@ from warnings import filterwarnings
import gc
import os
+from matplotlib.cbook import boxplot_stats
from matplotlib.colors import Normalize
from mplhep import style
from scipy.optimize import curve_fit
@@ -63,6 +64,7 @@ def _show_report(report_info):
print("{} - VFAT analyzed.".format(info["analyzed vfat"]))
print("{} - empty channels found.".format(info["empty channels"]))
print("{} - channels failed fitting.".format(info["fit-failed channels"]))
+ print("{} - noisy channels.".format(info["noisy channels"]))
print("------------------------------------------------------------")
@@ -121,12 +123,23 @@ def _analyze_vfat(vfat_dir):
"threshold": mean,
"enc": sigma,
"empty": status == None,
+ "noisy": None,
}
)
return pd.DataFrame.from_records(values)
+def _filter_noisy_channels(enc_data):
+ """Routine to filter noisy channels based on the outliers points defined by the interquartile range.
+ Only upper fliers are taken.
+
+ :param pandas.core.series.Series enc_data: Pandas series with the noise of the channels for a VFAT
+ """
+ stats = boxplot_stats(enc_data)[0]
+ return (enc_data.isin(stats["fliers"])) & (enc_data.gt(stats["med"]))
+
+
def _scurve_summary_plots(oh_dirs, graph_name, units=Units.DAC_UNITS, calibration_df=None):
"""Routine to produce a summary 2D plots of the VFATs for an OH
@@ -419,6 +432,7 @@ def create_configuration(
"threshold": None,
"enc": None,
"empty": None,
+ "noisy": None,
},
)
@@ -429,6 +443,7 @@ def create_configuration(
analyzed_oh = 0
empty_channels = 0
fit_failed_channels = 0
+ noisy_channels = 0
print("Analyzing", input_file_name)
input_file = uproot.open(input_file_name)
@@ -454,6 +469,11 @@ def create_configuration(
# Merge results into a single DataFrame
final_data = pd.concat([r.get() for r in results], ignore_index=True)
+ # Noisy channels are filtered
+ final_data["noisy"] = final_data.groupby(["fed", "slot", "oh", "vfat"])[
+ "enc"
+ ].transform(_filter_noisy_channels)
+
# Data are changed to desired units
calibration_df = None
if units == Units.FC:
@@ -583,6 +603,13 @@ def create_configuration(
].drop_duplicates(["fed", "slot", "oh", "vfat", "channel"])
)
+ # To count how many noisy channels exist
+ noisy_channels = len(
+ final_data.loc[(final_data["noisy"] == True)].drop_duplicates(
+ ["fed", "slot", "oh", "vfat", "channel"]
+ )
+ )
+
report_info += [
{
"filename": input_file_name,
@@ -590,6 +617,7 @@ def create_configuration(
"analyzed vfat": analyzed_vfat,
"empty channels": empty_channels,
"fit-failed channels": fit_failed_channels,
+ "noisy channels": noisy_channels,
}
]
--
GitLab