Add evaluation script

scripts/evaluation/classify_test_runs.py

@@ -0,0 +1,69 @@
+''' This script reads a journal file and filters out lines that begin with 'Created' or 'Opened' or <!> and then reads the remaining lines containing object in the RUNNING state into a pandas dataframe.
+    The script then groups the rows based on closeness of timestamps (2 minutes) and creates new journal files for each group.
+'''
+
+import pandas as pd
+import io
+import re
+
+filename = 'xxx.jnl'
+
+lines=[]
+#Filter lines that begin with 'Created' or 'Opened' or <!>
+with open(filename) as f:
+    for line in f:
+        if "Created" in line:
+            continue
+        if "Opened" in line:
+            continue
+        if "<!>" in line:
+            continue
+        lines.append(line)
+
+lines = [line.replace(';\n', '') for line in lines]
+lines = [re.sub(r'([0-9]+\.[0-9]+)\:\s[0-9]+\.[0-9]+', r'\1', line) for line in lines]
+buffer = io.StringIO("\n".join(lines))
+df=pd.read_csv(buffer, sep=';', names=['timestamp','client_id','ip','x','y','z','heading_rad','speed_long_m_s','speed_lat_m_s','accel_long_m_s2','accel_lat_m_s2','drive_direction','obj_state','arm_readiness','error_status'])
+df.drop_duplicates(subset=['timestamp'], keep='first', inplace=True)
+df['timestamp'] = pd.to_datetime(df['timestamp'],unit='s')
+df=df[df.obj_state.isin(['RUNNING'])]
+
+
+#Group the rows based on closeness of timestamps (2 minutes)
+df['Group'] = (df['timestamp'].diff().dt.total_seconds() > 2*60).cumsum()
+
+# print start and end timestamp of each group as well as count and duration in time
+print(df.groupby('Group').agg({'timestamp': ['first', 'last', 'count', lambda x: x.max() - x.min()]}))
+
+# create new journal files for each group
+for name, group in df.groupby('Group'):
+    group.to_csv(f'group_{name}.jnl', sep=';', index=False)
+
+
+import matplotlib.pyplot as plt
+import seaborn as sns
+import numpy as np
+
+# plot each group sequentially using seaborn with the title of each group (commented out for now)
+
+'''
+for name, group in df.groupby('Group'):
+    #plot the x and y coordinates of each object showing heading with an arrow
+    plt.figure()
+    sns.scatterplot(x='x', y='y', hue='client_id', palette="muted", data=group)
+    for i, row in group.iterrows():
+        plt.arrow(row['x'], row['y'], np.cos(row['heading_rad']), np.sin(row['heading_rad']), head_width=0.1, head_length=0.1, fc='k', ec='k')
+    plt.title(f'Group {name}')
+    plt.show()
+
+    # plot longitudinal speed for each of the 6 objects in the group
+    plt.figure()
+    sns.lineplot(x='timestamp', y='speed_long_m_s', hue='client_id', palette="muted", data=group)
+    plt.title(f'Group {name}')
+    plt.show()
+    # plot lateral speed for each of the 6 objects in the group
+    plt.figure()
+    sns.lineplot(x='timestamp', y='speed_lat_m_s', hue='client_id', palette="muted", data=group)
+    plt.title(f'Group {name}')
+    plt.show()
+'''