Updating scatterplot scripts to handle larger scales

src/scripts/counter_scatterplot.py

@@ -14,13 +14,6 @@
 
 def shorten_name(name):
     tmp = name
-    # remove arguments in function name, if they exist
-    leftparen = tmp.find('(')
-    rightparen = tmp.rfind(')')
-    if leftparen > 0 and rightparen > 0:
-        tokens1 = tmp.split('(')
-        tokens2 = tmp.split(')')
-        tmp = tokens1[0] + '()' + tokens2[len(tokens2)-1]
     # otherwise, just take the first 64 characters.
     short = (tmp[:77] + '...') if len(tmp) > 77 else tmp
     return short.replace('_', '$\_$')
@@ -34,21 +27,23 @@ def shorten_name(name):
             index = index + 1
             if len(row) == 3 and not row[0].strip().startswith("#"):
                 try:
-                    mytup = (float(row[0]),float(row[1]))
+                    mytup = (float(row[0])/1000000000.0,float(row[1]))
                 except ValueError as e:
                     print(index, " Bad row: ", row)
                     continue
                 if row[2] not in dictionary:
                     dictionary[row[2]] = [mytup]
                 else:
                     dictionary[row[2]].append(mytup)
+            if (index % 100000 == 0):
+                print (index, 'rows parsed...', end='\r', flush=True)
         print ("Parsed", index, "samples")
 
 #resize the figure
 # Get current size
 fig_size = pl.rcParams["figure.figsize"]
 # Set figure width to 12 and height to 9
-fig_size[0] = 8
+fig_size[0] = 16
 fig_size[1] = 4
 pl.rcParams["figure.figsize"] = fig_size
 
@@ -65,9 +60,12 @@ def shorten_name(name):
     values = np.array([x[1] for x in dictionary[key]])
     name = shorten_name(key)
     pl.title(name);
-    #pl.semilogy(timestamps, values, color=mycolor[index-1], marker=mymark[index-1], linestyle=' ', label=name)
-    #pl.plot(timestamps, values, color=mycolor[index-1], marker=mymark[index-1], linestyle=' ', label=name)
-    pl.plot(timestamps, values, marker='o', color='blue')
+    marker = '1'
+    extension = '.pdf'
+    if len(timestamps) > 1000000:
+        marker = ','
+        extension = '.png'
+    pl.plot(timestamps, values, marker=marker, color='blue', linestyle=' ', label=name)
     pl.draw()
     axes.set_autoscale_on(True) # enable autoscale
     axes.autoscale_view(True,True,True)
@@ -78,5 +76,5 @@ def shorten_name(name):
     pl.tight_layout()
     name = name.replace(" ", "_")
     name = ''.join([c for c in name if c in safechars])
-    pl.savefig(name+".png")
+    pl.savefig(name+extension)
     pl.close()

src/scripts/task_scatterplot.py

@@ -47,25 +47,28 @@ def shorten_name(name):
 # Get current size
 fig_size = pl.rcParams["figure.figsize"]
 # Set figure width to 12 and height to 9
-fig_size[0] = 12
-fig_size[1] = 9
+fig_size[0] = 16
+fig_size[1] = 10
 pl.rcParams["figure.figsize"] = fig_size
 
 mymark = ('o', 'v', '^', '<', '>', '8', 's', 'p', '*', 'h', 'H', 'D', 'd')
-mycolor = ('blue', 'green', 'red', 'cyan', 'magenta', 'yellow', 'black', 'darkblue', 'darkgreen', 'darkred')
+mycolor = ('blue', 'green', 'red', 'royalblue', 'darkmagenta', 'darkslategrey', 'black', 'darkblue', 'darkgreen', 'darkred')
+#mycolor = iter([pl.cm.Dark2(i) for i in range(8)])
 
 index = 0
 numplots = min(len(dictionary), 10)
 for key in sorted(dictionary, key=lambda key: len(dictionary[key]), reverse=True):
     index = index + 1
     print ("Plotting", key)
     axes = pl.subplot(((numplots+1)/2), 2, index)
-    timestamps = np.array([x[0] for x in dictionary[key]])
-    values = np.array([x[1] for x in dictionary[key]])
+    timestamps = np.array([x[0]/1000000000 for x in dictionary[key]])
+    values = np.array([x[1]/1000 for x in dictionary[key]])
     name = shorten_name(key)
     axes.set_title(name);
+    pl.plot(timestamps, values, color=mycolor[index-1], marker='.', linestyle=' ', label=name)
+    #pl.plot(timestamps, values, color=mycolor[next(mycolor)], marker='.', linestyle=' ', label=name)
     #pl.plot(timestamps, values, color=mycolor[index-1], marker=mymark[index-1], linestyle=' ', label=name)
-    pl.semilogy(timestamps, values, color=mycolor[index-1], marker=mymark[index-1], linestyle=' ', label=name)
+    #pl.semilogy(timestamps, values, color=mycolor[index-1], marker=mymark[index-1], linestyle=' ', label=name)
     pl.draw()
     axes.set_autoscale_on(True) # enable autoscale
     axes.autoscale_view(True,True,True)