From 1a0e4599c530c197a73dabb919859d8f37d60b7a Mon Sep 17 00:00:00 2001
From: Ole Laursen <olau@iola.dk>
Date: Thu, 16 Feb 2017 12:23:45 +0000
Subject: [PATCH] Adjust the way authors with unknown countries are counted and
 improve the explanation of how the numbers have come to be  - Legacy-Id:
 12858

---
 ietf/stats/views.py                              | 16 ++++++++++++++++
 .../stats/document_stats_author_continent.html   |  9 ++++-----
 .../stats/document_stats_author_country.html     | 12 +++++++-----
 3 files changed, 27 insertions(+), 10 deletions(-)

diff --git a/ietf/stats/views.py b/ietf/stats/views.py
index 68e3f29d0..3466e329a 100644
--- a/ietf/stats/views.py
+++ b/ietf/stats/views.py
@@ -433,6 +433,14 @@ def document_stats(request, stats_type=None):
                 if c and c.in_eu:
                     bins[eu_name].append(name)
 
+            # remove from the unknown bin all authors with a known country
+            all_known = set(n for b, names in bins.iteritems() if b for n in names)
+            unknown = []
+            for name in bins[""]:
+                if name not in all_known:
+                    unknown.append(name)
+            bins[""] = unknown
+
             series_data = []
             for country, names in sorted(bins.iteritems(), key=lambda t: t[0].lower()):
                 percentage = len(names) * 100.0 / total_persons
@@ -470,6 +478,14 @@ def document_stats(request, stats_type=None):
                 continent_name = country_to_continent.get(country_name, "")
                 bins[continent_name].append(name)
 
+            # remove from the unknown bin all authors with a known continent
+            all_known = set(n for b, names in bins.iteritems() if b for n in names)
+            unknown = []
+            for name in bins[""]:
+                if name not in all_known:
+                    unknown.append(name)
+            bins[""] = unknown
+
             series_data = []
             for continent, names in sorted(bins.iteritems(), key=lambda t: t[0].lower()):
                 percentage = len(names) * 100.0 / total_persons
diff --git a/ietf/templates/stats/document_stats_author_continent.html b/ietf/templates/stats/document_stats_author_continent.html
index d0327bf3a..b17d49552 100644
--- a/ietf/templates/stats/document_stats_author_continent.html
+++ b/ietf/templates/stats/document_stats_author_continent.html
@@ -58,8 +58,7 @@
   </tbody>
 </table>
 
-<p>The country information for an author can vary between documents,
-  so the sum of the rows in the table can be more than 100%. This
-  is especially true for the row with unknown continent information -
-  many authors may have one or more author entries with an
-  unrecognized country.</p>
+<p>The statistics are based entirely on the author addresses provided
+  in each draft. Since this varies across documents, a travelling
+  author may be counted in more than country, making the total sum
+  more than 100%.</p>
diff --git a/ietf/templates/stats/document_stats_author_country.html b/ietf/templates/stats/document_stats_author_country.html
index 1b8911c4c..99c6d0152 100644
--- a/ietf/templates/stats/document_stats_author_country.html
+++ b/ietf/templates/stats/document_stats_author_country.html
@@ -58,11 +58,13 @@
   </tbody>
 </table>
 
-<p>The country information for an author can vary between documents,
-  so the sum of multiple rows in the table can be more than 100%. This
-  is especially true for the row with unknown country information -
-  many authors may have one or more author entries with an
-  unrecognized country.</p>
+<p>The statistics are based entirely on the author addresses provided
+  in each draft. Since this varies across documents, a travelling
+  author may be counted in more than country, making the total sum
+  more than 100%.</p>
+
+<p>In case no country information is found for an author in the time
+  period, the author is counted as (unknown).</p>
 
 <p>An author is counted in EU if the country is a member of the EU
   now, even if that was not the case at publication.