From f6f2b3c495cce8064b2fd1afb720be4fe0e4f03d Mon Sep 17 00:00:00 2001
From: Sanj <b@pad.ma>
Date: Fri, 8 Jun 2012 13:33:35 +0530
Subject: [PATCH 1/3] fuzzystops

---
 chaloBEST/mumbai/views.py           | 31 ++++++++++++++++++++++++++++-
 chaloBEST/templates/fuzzystops.html | 14 +++++++++++++
 chaloBEST/urls.py                   |  1 +
 3 files changed, 45 insertions(+), 1 deletion(-)
 create mode 100644 chaloBEST/templates/fuzzystops.html

diff --git a/chaloBEST/mumbai/views.py b/chaloBEST/mumbai/views.py
index 2ee5836..d9289ce 100644
--- a/chaloBEST/mumbai/views.py
+++ b/chaloBEST/mumbai/views.py
@@ -3,6 +3,7 @@ from models import *
 from django.shortcuts import render_to_response, get_object_or_404
 from django.template import RequestContext
 from django.contrib.auth.decorators import login_required
+from fuzzywuzzy import process as fuzzprocess
 
 def index(request):
     return render_to_response("index.html", {})
@@ -103,4 +104,32 @@ def stats(request):
 
 
 
-
+@login_required
+def fuzzystops(request):
+#    import pdb
+    froms_arr = []
+    tos_arr = []
+    for unr in UniqueRoute.objects.all():
+        s1 = unr.from_stop.name.lower()
+        s2 = unr.from_stop_txt.lower()
+        from_ratio = fuzzprocess.ratio(s1,s2)
+        if from_ratio < 50:
+            froms_arr.append(
+                (unr, from_ratio,)
+            ) 
+        s3 = unr.to_stop.name.lower()
+        s4 = unr.to_stop_txt.lower()
+        to_ratio = fuzzprocess.ratio(s3,s4)
+        if to_ratio < 50:
+            tos_arr.append(
+                (unr, to_ratio,)
+            )
+            
+    froms_arr.sort(key=lambda item: item[1])
+    tos_arr.sort(key=lambda item: item[1])
+    context = RequestContext(request, {
+        'fuzzy_froms': [item[0] for item in froms_arr],
+        'fuzzy_tos': [item[0] for item in tos_arr]
+    })
+#    pdb.set_trace()
+    return render_to_response("fuzzystops.html", context)
diff --git a/chaloBEST/templates/fuzzystops.html b/chaloBEST/templates/fuzzystops.html
new file mode 100644
index 0000000..0388497
--- /dev/null
+++ b/chaloBEST/templates/fuzzystops.html
@@ -0,0 +1,14 @@
+<h2> Likely Mismatched From stops: </h2>
+{% for f in fuzzy_froms %}
+
+    <a href="/admin/mumbai/uniqueroute/{{ f.id }}">{{ f.route.alias}}</a> - {{ f.from_stop.name }} - {{ f.from_stop_txt }} <br />
+
+{% endfor %}
+
+<h2> Likely mismatched to stops: </h2> 
+
+{% for t in fuzzy_tos %}
+
+    <a href="/admin/mumbai/uniqueroute/{{ t.id }}">{{ t.route.alias}}</a> - {{ t.from_stop.name }} - {{ t.from_stop_txt }} <br />
+
+{% endfor %}
diff --git a/chaloBEST/urls.py b/chaloBEST/urls.py
index ab4f134..06c26f0 100644
--- a/chaloBEST/urls.py
+++ b/chaloBEST/urls.py
@@ -31,6 +31,7 @@ urlpatterns = patterns('',
     #(r'^grappelli/', include('grappelli.urls')),
     # Uncomment the next line to enable the admin:
     (r'^admin/', include(admin.site.urls)),
+    (r'^fuzzystops/$', 'mumbai.views.fuzzystops'),
     
 )
 

From 40c6f4fab3d91f613e74b4891e5eb846da875127 Mon Sep 17 00:00:00 2001
From: Sanj <b@pad.ma>
Date: Fri, 8 Jun 2012 13:48:45 +0530
Subject: [PATCH 2/3] silly template error in fuzzy template

---
 chaloBEST/templates/fuzzystops.html | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/chaloBEST/templates/fuzzystops.html b/chaloBEST/templates/fuzzystops.html
index 0388497..8b4171f 100644
--- a/chaloBEST/templates/fuzzystops.html
+++ b/chaloBEST/templates/fuzzystops.html
@@ -9,6 +9,6 @@
 
 {% for t in fuzzy_tos %}
 
-    <a href="/admin/mumbai/uniqueroute/{{ t.id }}">{{ t.route.alias}}</a> - {{ t.from_stop.name }} - {{ t.from_stop_txt }} <br />
+    <a href="/admin/mumbai/uniqueroute/{{ t.id }}">{{ t.route.alias}}</a> - {{ t.to_stop.name }} - {{ t.to_stop_txt }} <br />
 
 {% endfor %}

From cff4beba26fbd7fe086a66ae15be1d9866d6fdb4 Mon Sep 17 00:00:00 2001
From: Sanj <b@pad.ma>
Date: Fri, 8 Jun 2012 14:11:02 +0530
Subject: [PATCH 3/3] increase fuzzy threshhold

---
 chaloBEST/mumbai/views.py | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/chaloBEST/mumbai/views.py b/chaloBEST/mumbai/views.py
index d9289ce..a04e914 100644
--- a/chaloBEST/mumbai/views.py
+++ b/chaloBEST/mumbai/views.py
@@ -113,14 +113,14 @@ def fuzzystops(request):
         s1 = unr.from_stop.name.lower()
         s2 = unr.from_stop_txt.lower()
         from_ratio = fuzzprocess.ratio(s1,s2)
-        if from_ratio < 50:
+        if from_ratio < 70:
             froms_arr.append(
                 (unr, from_ratio,)
             ) 
         s3 = unr.to_stop.name.lower()
         s4 = unr.to_stop_txt.lower()
         to_ratio = fuzzprocess.ratio(s3,s4)
-        if to_ratio < 50:
+        if to_ratio < 70:
             tos_arr.append(
                 (unr, to_ratio,)
             )