From f7d3245f925fd7cd5115e2064d2fff07a0475f8c Mon Sep 17 00:00:00 2001
From: Michael Reuscher <michael.reuscher@desy.de>
Date: Wed, 26 Jul 2023 11:47:39 +0200
Subject: [PATCH] statistics plot 1.1
---
visualizer/visualizer/DataFrameCreator.py | 14 ++++++++-
visualizer/visualizer/SeabornPlotter.py | 38 ++++++++++++++++++++---
visualizer/visualizer/main.py | 20 ++++++------
3 files changed, 57 insertions(+), 15 deletions(-)
diff --git a/visualizer/visualizer/DataFrameCreator.py b/visualizer/visualizer/DataFrameCreator.py
index f227893..cc518f1 100644
--- a/visualizer/visualizer/DataFrameCreator.py
+++ b/visualizer/visualizer/DataFrameCreator.py
@@ -1,7 +1,6 @@
import os
import pandas as pd
import re
-from collections import OrderedDict
class DataFrameCreator:
@@ -53,6 +52,16 @@ class DataFrameCreator:
return result
+ @staticmethod
+ def _extract_avg(data):
+ result = {}
+ match = re.search(r"(\d+\.\d+)", data)
+ if match:
+ result["avg"] = float(match.group())
+ else:
+ result["avg"] = 0.0
+ return result
+
def read_various_data(self, root, filename):
dataframe, dataframe_type, dataframe_name = self._set_dataframe(root, filename)
if dataframe_type in self.v_dfs:
@@ -79,6 +88,9 @@ class DataFrameCreator:
if data.startswith("worker") or data.startswith("posix"):
data_dict = self._extract_statistics(data, filename)
data_list.append(data_dict)
+ elif data.startswith("average"):
+ data_dict = self._extract_avg(data)
+ data_list.append(data_dict)
dataframe = pd.DataFrame(data_list)
dataframe_type = filename
diff --git a/visualizer/visualizer/SeabornPlotter.py b/visualizer/visualizer/SeabornPlotter.py
index ef44dc7..0509ac5 100644
--- a/visualizer/visualizer/SeabornPlotter.py
+++ b/visualizer/visualizer/SeabornPlotter.py
@@ -1,5 +1,6 @@
import numpy as np
import matplotlib.pyplot as plt
+import pandas as pd
import seaborn as sns
from matplotlib.offsetbox import OffsetImage, AnnotationBbox
import os
@@ -11,6 +12,15 @@ class SeabornPlotter:
self.dataframe = dataframe
self.name = name
+ @staticmethod
+ def format_speed(speed_gibs):
+ # Convert speed from GiB/s to MiB/s
+ speed_mibs = speed_gibs * 1024
+
+ # Round the speed to two decimal places
+ speed_mibs_rounded = round(speed_mibs, 2)
+ return speed_mibs_rounded
+
@staticmethod
def _add_logo(logo_path, logo_size, logo_x, logo_y):
current_directory = os.path.dirname(os.path.abspath(__file__))
@@ -67,9 +77,25 @@ class SeabornPlotter:
plt.savefig(title + ".svg", format='svg')
plt.show()
+ def plot_avg_speed(self, speed_dict):
+ data_list = list(speed_dict.items())
+ df = pd.DataFrame(data_list, columns=["Data_set", "MiB/s"])
+
+ # Plot erstellen
+ plt.figure(figsize=(12, 6))
+ sns.lineplot(x=df["Data_set"], y=df["MiB/s"], marker='o')
+ plt.xticks(rotation=90) # Rotiere die x-Achsenbeschriftungen für bessere Lesbarkeit
+ self._add_logo("desy_logo.png", logo_size=0.1, logo_x=1.065, logo_y=1.07)
+ plt.xlabel('Data_Set')
+ plt.yscale('log')
+ plt.ylabel('MiB/s')
+ plt.title('AVG Speed')
+ plt.tight_layout()
+ plt.show()
+
def plot_statistics(self):
dataframes = self.dataframe
-
+ avg_speed_dict = {}
# Sort the dataframes based on the 'size' column in ascending order
sorted_dataframes = sorted(dataframes.items(), key=lambda x: x[1]['size'].min())
@@ -78,20 +104,24 @@ class SeabornPlotter:
cols = (num_plots + 1) // rows # Number of columns in the grid
# Subplots creation
- fig, axes = plt.subplots(rows, cols, figsize=(180, 120), constrained_layout=True)
- fig.suptitle('Boxplot of Statistical Measures for Each Group', fontsize=14)
+ fig, axes = plt.subplots(rows, cols, figsize=(180, 120), constrained_layout=True)
+ fig.suptitle('Statistical Measures for Each Data_Set', fontsize=14)
# For each DataFrame, create a boxplot and place it into the corresponding subplot
for idx, (df_key, df) in enumerate(sorted_dataframes):
- sns.boxplot(data=df, ax=axes[idx // cols, idx % cols], palette='pastel')
+ sns.boxplot(data=df, order=['min', 'max', 'mean', 'med', '10%', '90%'], ax=axes[idx // cols, idx % cols], palette='pastel')
axes[idx // cols, idx % cols].set_yscale('log')
axes[idx // cols, idx % cols].set_ylabel('Time (Y-Axis)')
axes[idx // cols, idx % cols].set_title(f'Group {df_key}', fontsize=12)
axes[idx // cols, idx % cols].grid(True)
+ first_valid_index = df['avg'].first_valid_index()
+ speed_value = df.loc[first_valid_index, 'avg']
+ avg_speed_dict[df_key] = self.format_speed(speed_value)
# Hide empty subplots, if any
for idx in range(num_plots, rows * cols):
fig.delaxes(axes.flatten()[idx])
plt.show()
+ self.plot_avg_speed(avg_speed_dict)
diff --git a/visualizer/visualizer/main.py b/visualizer/visualizer/main.py
index f204c88..d6622fc 100644
--- a/visualizer/visualizer/main.py
+++ b/visualizer/visualizer/main.py
@@ -23,21 +23,21 @@ def main():
s_plotter.plot_statistics()
# Plot Data
- #for frame in v_dataframes.keys():
- #if not v_dataframes[frame].empty:
+ for frame in v_dataframes.keys():
+ if not v_dataframes[frame].empty:
# Setup Plotter
- #frame_name = frame
- #v_plotter = SeabornPlotter(v_dataframes[frame], frame_name)
+ frame_name = frame
+ v_plotter = SeabornPlotter(v_dataframes[frame], frame_name)
# Plotter run
- #v_plotter.plot_histogram_various()
+ v_plotter.plot_histogram_various()
- #for frame in l_dataframes.keys():
- #if not v_dataframes[frame].empty:
+ for frame in l_dataframes.keys():
+ if not v_dataframes[frame].empty:
# Setup Plotter
- #frame_name = frame
- #l_plotter = SeabornPlotter(l_dataframes[frame], frame_name)
+ frame_name = frame
+ l_plotter = SeabornPlotter(l_dataframes[frame], frame_name)
# Plotter run
- #l_plotter.plot_histogram_long()
+ l_plotter.plot_histogram_long()
if __name__ == "__main__":
--
GitLab