Message distribution plot improvement

src/mesh_analysis/waku_message_log_analyzer.py

@@ -3,6 +3,10 @@
 import base64
 import logging
 import pandas as pd
+import matplotlib.pyplot as plt
+import matplotlib.patheffects as path_effects
+import numpy as np
+import seaborn as sns
 from concurrent.futures import ProcessPoolExecutor, as_completed
 from pathlib import Path
 from typing import List, Dict
@@ -12,6 +16,7 @@
 from src.mesh_analysis.readers.file_reader import FileReader
 from src.mesh_analysis.readers.victoria_reader import VictoriaReader
 from src.mesh_analysis.tracers.waku_tracer import WakuTracer
+from src.plotting.utils import add_boxplot_stat_labels
 from src.utils import file_utils, log_utils, path_utils, list_utils
 
 logger = logging.getLogger(__name__)
@@ -324,12 +329,19 @@ def analyze_message_timestamps(self, time_difference_threshold: int):
                 logger.info(f'{file}: {jump[0]} to {jump[1]} -> {jump[2]}')
 
 
-    def plot_message_distribution(self):
-        import matplotlib.pyplot as plt
-        import seaborn as sns
+    def plot_message_distribution(self, received_summary_path: Path, plot_title: str, dump_path: Path) -> Result[None, str]:
+        """
+        Note that this function assumes that analyze_message_logs has been called, since timestamps will be checked
+        from logs.
+        """
+        if not received_summary_path.exists():
+            error = f'Received summary file {received_summary_path} does not exist'
+            logger.error(error)
+            return Err(error)
+
         sns.set_theme()
 
-        df = pd.read_csv('local_data/mixed_enviroment/summary/received.csv', parse_dates=['timestamp'])
+        df = pd.read_csv(received_summary_path, parse_dates=['timestamp'])
         df.set_index(['shard', 'msg_hash', 'timestamp'], inplace=True)
 
         time_ranges = df.groupby(level='msg_hash').apply(
@@ -340,10 +352,32 @@ def plot_message_distribution(self):
         time_ranges_df = time_ranges.reset_index(name='time_to_reach')
 
         plt.figure(figsize=(12, 6))
-        sns.boxplot(x='time_to_reach', data=time_ranges_df, color='skyblue')
+        ax = sns.boxplot(x='time_to_reach', data=time_ranges_df, color='skyblue', whis=(0,100))
+
+        add_boxplot_stat_labels(ax, value_type="min")
+        add_boxplot_stat_labels(ax, value_type="max")
+        add_boxplot_stat_labels(ax, value_type="median")
+
+        q1 = np.percentile(time_ranges_df['time_to_reach'], 25)
+        q3 = np.percentile(time_ranges_df['time_to_reach'], 75)
+
+        text = ax.text(y=-0.1, x=q1, s=f'{q1:.3f}', ha='center', va='center',
+                       fontweight='bold', color='white', size=10)
+        text.set_path_effects([
+            path_effects.Stroke(linewidth=3, foreground=ax.get_lines()[0].get_color()),
+            path_effects.Normal(),
+        ])
+        text = ax.text(y=-0.1, x=q3, s=f'{q3:.3f}', ha='center', va='center',
+                       fontweight='bold', color='white', size=10)
+        text.set_path_effects([
+            path_effects.Stroke(linewidth=3, foreground=ax.get_lines()[0].get_color()),
+            path_effects.Normal(),
+        ])
 
         plt.xlabel('Time to Reach All Nodes (seconds)')
-        plt.title('210 Nodes - 1msg/s - 1KB - 600 messages \n Message time distribution')
+        plt.title(plot_title)
 
-        plt.savefig("distribution-mixed")
+        plt.savefig(dump_path)
         plt.show()
+
+        return Ok(None)

src/plotting/plotter.py

@@ -4,13 +4,13 @@
 import pandas as pd
 import seaborn as sns
 import matplotlib.pyplot as plt
-import matplotlib.patheffects as path_effects
 from typing import List, Dict
 from matplotlib import ticker
 
 # Project Imports
 from src.data.data_handler import DataHandler
 from src.data.data_file_handler import DataFileHandler
+from src.plotting.utils import add_boxplot_stat_labels
 
 logger = logging.getLogger(__name__)
 sns.set_theme()
@@ -67,53 +67,24 @@ def _add_subplot_df_to_axs(self, df: pd.DataFrame, index: int, axs: np.ndarray,
         box_plot.xaxis.set_tick_params(rotation=45)
         box_plot.legend(loc='upper right', bbox_to_anchor=(1, 1))
 
-        self._add_stat_labels(box_plot)
+        result = add_boxplot_stat_labels(box_plot)
+        if result.is_err():
+            logger.error(result.err_value)
+
         show_min_max = plot_specs.get("show_min_max", False)
         if show_min_max:
-            self._add_stat_labels(box_plot, value_type="min")
-            self._add_stat_labels(box_plot, value_type="max")
+            result = add_boxplot_stat_labels(box_plot, value_type="min")
+            if result.is_err():
+                logger.error(result.err_value)
+
+            result = add_boxplot_stat_labels(box_plot, value_type="max")
+            if result.is_err():
+                logger.error(result.err_value)
 
     def _create_subplot_paths_group(self, plot_specs: Dict) -> List:
         subplot_path = [[f"{folder}{data}" for folder in plot_specs["folder"]] for data in
                         plot_specs["data"]]
 
         return subplot_path
 
-    def _add_stat_labels(self, ax: plt.Axes, fmt: str = ".3f", value_type: str = "median") -> None:
-        # Refactor from https://stackoverflow.com/a/63295846
-        """
-        Add text labels to the median, minimum, or maximum lines of a seaborn boxplot.
-
-        Args:
-            ax: plt.Axes, e.g., the return value of sns.boxplot()
-            fmt: Format string for the value (e.g., min/max/median).
-            value_type: The type of value to label. Can be 'median', 'min', or 'max'.
-        """
-        lines = ax.get_lines()
-        boxes = [c for c in ax.get_children() if "Patch" in str(c)]  # Get box patches
-        start = 4
-        if not boxes:  # seaborn v0.13 or above (no patches => need to shift index)
-            boxes = [c for c in ax.get_lines() if len(c.get_xdata()) == 5]
-            start += 1
-        lines_per_box = len(lines) // len(boxes)
-
-        if value_type == "median":
-            line_idx = start
-        elif value_type == "min":
-            line_idx = start - 2  # min line comes 2 positions before the median
-        elif value_type == "max":
-            line_idx = start - 1  # max line comes 1 position before the median
-        else:
-            raise ValueError("Invalid value_type. Must be 'min', 'max', or 'median'.")
-
-        for value_line in lines[line_idx::lines_per_box]:
-            x, y = (data.mean() for data in value_line.get_data())
-            # choose value depending on horizontal or vertical plot orientation
-            value = x if len(set(value_line.get_xdata())) == 1 else y
-            text = ax.text(x, y, f'{value / 1000:{fmt}}', ha='center', va='center',
-                           fontweight='bold', color='white', size=10)
-            # create colored border around white text for contrast
-            text.set_path_effects([
-                path_effects.Stroke(linewidth=3, foreground=value_line.get_color()),
-                path_effects.Normal(),
-            ])
+

src/plotting/utils.py

@@ -0,0 +1,48 @@
+# Python Imports
+import matplotlib.pyplot as plt
+import matplotlib.patheffects as path_effects
+from result import Err, Result, Ok
+
+# Project Imports
+
+
+def add_boxplot_stat_labels(ax: plt.Axes, fmt: str = ".3f", value_type: str = "median") -> Result[None, str]:
+    # Refactor from https://stackoverflow.com/a/63295846
+    """
+    Add text labels to the median, minimum, or maximum lines of a seaborn boxplot.
+
+    Args:
+        ax: plt.Axes, e.g., the return value of sns.boxplot()
+        fmt: Format string for the value (e.g., min/max/median).
+        value_type: The type of value to label. Can be 'median', 'min', or 'max'.
+    """
+    lines = ax.get_lines()
+    boxes = [c for c in ax.get_children() if "Patch" in str(c)]  # Get box patches
+    start = 4
+    if not boxes:  # seaborn v0.13 or above (no patches => need to shift index)
+        boxes = [c for c in ax.get_lines() if len(c.get_xdata()) == 5]
+        start += 1
+    lines_per_box = len(lines) // len(boxes)
+
+    if value_type == "median":
+        line_idx = start
+    elif value_type == "min":
+        line_idx = start - 2  # min line comes 2 positions before the median
+    elif value_type == "max":
+        line_idx = start - 1  # max line comes 1 position before the median
+    else:
+        return Err("Invalid value_type. Must be 'min', 'max', or 'median'.")
+
+    for value_line in lines[line_idx::lines_per_box]:
+        x, y = (data.mean() for data in value_line.get_data())
+        # choose value depending on horizontal or vertical plot orientation
+        value = x if len(set(value_line.get_xdata())) == 1 else y
+        text = ax.text(x, y, f'{value / 1000:{fmt}}', ha='center', va='center',
+                       fontweight='bold', color='white', size=10)
+        # create colored border around white text for contrast
+        text.set_path_effects([
+            path_effects.Stroke(linewidth=3, foreground=value_line.get_color()),
+            path_effects.Normal(),
+        ])
+
+    return Ok(None)