Added data/logAnalysis for extracting data from log files

data/logAnalysis.py

+# Open file
+# Read line by line those containing ^.* main [^\t]*
+# as W\ (a\ =\ v{;})*
+# To store as a line
+# Average (as float) all those lines with the same W
+
+import re
+import pandas as pd
+import pysake.data.tableDisplay as tdis
+
+def parseLogfile( filename, cols ):
+    """
+    Parse a log file and returns a dataframe with average attribute values aggregated by modality
+    Parameters:
+    :filename -- the name (path) of the log file
+    :cols -- the list of column names of the resulting table
+    """
+    logfile = open( filename, 'r' )
+    entries = []
+    for line in logfile:        
+        # Unsure it will always be DEBUG...
+        twosides = re.split(r'^.*DEBUG main [^\s]*\s+',line.strip())
+        if len(twosides) > 1:
+            tug = re.split( r"\s", twosides[1].strip(), maxsplit=1 )
+            listatt = re.findall( r'\s*\w+ = ([^;]*)[;]?', tug[1] )
+            entries += [ [tug[0]]+listatt ]
+    logfile.close()
+    df = pd.DataFrame( entries, columns=cols )
+    tdis.typeTableColumns( df )
+    df = df.groupby(cols[0]).mean()
+    return df

data/tableDisplay.py

@@ -142,5 +142,5 @@ def formatTableStandard( style, size=0 ):
         dict(selector="tr:nth-child("+str(size)+"n)", props=barbelow)
     ]
     style.set_table_styles( standardStyles )
-    style.format({'Size': '{:0.0f}','Success rate': '{:0.2f}','Incoherence': '{:0.2f}','Semantic precision': '{:0.2f}','Semantic F-measure': '{:0.2f}','Semantic recall': '{:0.2f}','Convergence': '{:0.0f}'})
+    style.format({'Size': '{:0.0f}','Success rate': '{:0.2f}','Incoherence': '{:0.2f}','Semantic precision': '{:0.2f}','Semantic F-measure': '{:0.2f}','Semantic recall': '{:0.2f}','Convergence': '{:0.0f}'} )
     return style