[tor-commits] [metrics-tasks/master] Tweak delber's relay-stats script a bit (#6329).

[karsten at torproject.org]

commit dd6329a8d82329033900fe95dd210123bfdb8bac
Author: Karsten Loesing <karsten.loesing at gmx.net>
Date:   Sun Jul 15 10:17:19 2012 +0200

    Tweak delber's relay-stats script a bit (#6329).
    
    - Include more information in the relays output, including Exit/Guard
      flag, country code, AS number, and AS name.
    - Add an option to filter relays by AS number.
    - Always print out consensus weights in relation to *all* running
      relays, not to the subset matching the given flags, countries, or
      ASes.  While it may be convenient to see fractions of "ASes when
      looking at Exit relays in .se", those numbers might be misleading if
      there are hardly any Exit relays in .se.  The better approach is to
      always compare the same consensus weights and also learn about the
      total consensus weight of the selection.  Do the latter by printing
      out total weights of the selection, unless all relays were selected.
      (See example below.)
    - Print out weights of relays/countries/ASes that didn't make it into
      the top-X.  (See example below.)
    
    $ ./tor-relays-stats.py as-sets 5 "Exit" "se"
      2.7014% ViaEuropa Sweden
      1.8749% Teknikbyran i Sverige AB
      0.5528% Bahnhof Internet AB
      0.5115% Foreningen for digitala fri- och rattigheter
      0.3397% Ownit Broadband AB
      0.2049% (10 others)
      6.1854% (total in selection)
---
 task-6329/tor-relays-stats.py |   53 +++++++++++++++++++++++++++++++++-------
 1 files changed, 43 insertions(+), 10 deletions(-)

diff --git a/task-6329/tor-relays-stats.py b/task-6329/tor-relays-stats.py
index cbb70ac..7efd623 100755
--- a/task-6329/tor-relays-stats.py
+++ b/task-6329/tor-relays-stats.py
@@ -29,7 +29,7 @@ class RelayStats(object):
             total_consensus_weight += relay['consensus_weight']
         return total_consensus_weight
 
-    def get_relays(self, flags=[], countries=''):
+    def get_relays(self, flags=[], countries='', as_sets=[]):
         relays = []
         for relay in self.data['relays']:
             if not relay['running']:
@@ -38,6 +38,8 @@ class RelayStats(object):
                 continue
             if countries and not relay.get('country', '') in countries:
                 continue
+            if as_sets and not relay.get('as_number', ' ') in as_sets:
+                continue
             relays.append(relay)
         return relays
 
@@ -54,9 +56,19 @@ class RelayStats(object):
 
         ranking = sorted(countries.iteritems(), key=operator.itemgetter(1))
         ranking.reverse()
-        total_consensus_weight = self.get_total_consensus_weight(relays)
+        total_consensus_weight = self.get_total_consensus_weight()
         for country, weight in ranking[:count]:
-            print "%3.2f%% %s" % (weight * 100.0 / total_consensus_weight, country)
+            print "%8.4f%% %s" % (weight * 100.0 / total_consensus_weight, country)
+        if len(ranking) > count:
+            other_consensus_weight = 0
+            for as_set, weight in ranking[count:]:
+                other_consensus_weight += weight
+            print "%8.4f%% (%d others)" % (other_consensus_weight * 100.0 / total_consensus_weight, len(ranking) - count)
+        selection_consensus_weight = 0
+        for as_set, weight in ranking:
+            selection_consensus_weight += weight
+        if selection_consensus_weight < total_consensus_weight:
+            print "%8.4f%% (total in selection)" % (selection_consensus_weight * 100.0 / total_consensus_weight)
 
     def output_as_sets(self, count='10', flags='', countries=''):
         count = int(count)
@@ -69,22 +81,43 @@ class RelayStats(object):
                 as_sets[as_set] = 0
             as_sets[as_set] += relay['consensus_weight']
 
-        total_consensus_weight = self.get_total_consensus_weight(relays)
+        total_consensus_weight = self.get_total_consensus_weight()
         ranking = sorted(as_sets.iteritems(), key=operator.itemgetter(1))
         ranking.reverse()
         for as_set, weight in ranking[:count]:
-            print "%3.4f%% %s" % (weight * 100.0 / total_consensus_weight, as_set)
-
-    def output_relays(self, count='10', flags='', countries=''):
+            print "%8.4f%% %s" % (weight * 100.0 / total_consensus_weight, as_set)
+        if len(ranking) > count:
+            other_consensus_weight = 0
+            for as_set, weight in ranking[count:]:
+                other_consensus_weight += weight
+            print "%8.4f%% (%d others)" % (other_consensus_weight * 100.0 / total_consensus_weight, len(ranking) - count)
+        selection_consensus_weight = 0
+        for as_set, weight in ranking:
+            selection_consensus_weight += weight
+        if selection_consensus_weight < total_consensus_weight:
+            print "%8.4f%% (total in selection)" % (selection_consensus_weight * 100.0 / total_consensus_weight)
+
+    def output_relays(self, count='10', flags='', countries='', as_sets=''):
         count = int(count)
         flags = flags.split()
-        relays = self.get_relays(flags, countries)
+        as_sets = as_sets.split()
+        relays = self.get_relays(flags, countries, as_sets)
 
         total_consensus_weight = self.get_total_consensus_weight()
         ranking = sorted(relays, key=operator.itemgetter('consensus_weight'))
         ranking.reverse()
+        selection_consensus_weight = 0
         for relay in ranking[:count]:
-            print "%3.4f%% %-20s %s" % (relay['consensus_weight'] * 100.0 / total_consensus_weight, relay['nickname'], relay['fingerprint'])
+            selection_consensus_weight += relay['consensus_weight']
+            print "%8.4f%% %-19s %-2s %-4s %-5s %s %-7s %s" % (relay['consensus_weight'] * 100.0 / total_consensus_weight, relay['nickname'], relay['fingerprint'], 'Exit' if 'Exit' in set(relay['flags']) else '', 'Guard' if 'Guard' in set(relay['flags']) else '', relay.get('country', ''), relay.get('as_number', ''), relay.get('as_name', ''))
+        if len(ranking) > count:
+            other_consensus_weight = 0
+            for relay in ranking[count:]:
+                other_consensus_weight += relay['consensus_weight']
+                selection_consensus_weight += relay['consensus_weight']
+            print "%8.4f%% (%d others)" % (other_consensus_weight * 100.0 / total_consensus_weight, len(ranking) - count)
+        if selection_consensus_weight < total_consensus_weight:
+            print "%8.4f%% (total in selection)" % (selection_consensus_weight * 100.0 / total_consensus_weight)
 
 OUTPUTS = {
   'countries': 'output_countries',
@@ -100,7 +133,7 @@ Where <output> is one of:
    relative percentage of the consensus in each countries
  - as-sets [COUNT] [FLAGS] [COUNTRIES]
    relative percentage of the consensus in each AS sets
- - relays [COUNT] [FLAGS] [COUNTRIES]
+ - relays [COUNT] [FLAGS] [COUNTRIES] [AS_SETS]
    list relays ranked by their place in the whole consensus
 
 Examples: