added weighted average - rScore now uses weighted average

rowers/dataprep.py

@@ -56,7 +56,7 @@ import sys
 
 import utils
 import datautils
-from utils import lbstoN,myqueue,is_ranking_piece
+from utils import lbstoN,myqueue,is_ranking_piece,wavg
 
 from timezonefinder import TimezoneFinder
 
@@ -2234,16 +2234,19 @@ def workout_rscore(w):
         df,row = getrowdata_db(id=w.id)
         df = clean_df_stats(df,workstrokesonly=False)
 
+    df['deltat'] = df['time'].diff()
     duration = df['time'].max()-df['time'].min()
     duration /= 1.0e3
-    pwr4 = df['power']**(4.0)
+    df['pwr4'] = df['power']**(4.0)
-    normp = (pwr4.mean())**(0.25)
+    pwr4mean = wavg(df,'pwr4','deltat')
+    pwrmean = wavg(df,'power','deltat')
+    normp = (pwr4mean)**(0.25)
     if not np.isnan(normp):
         ftp = float(r.ftp)
         if w.workouttype in ('water','coastal'):
             ftp = ftp*(100.-r.otwslack)/100.
 
-        intensityfactor = df['power'].mean()/float(ftp)
+        intensityfactor = pwrmean/float(ftp)
         intensityfactor = normp/float(ftp)
         tss = 100.*((duration*normp*intensityfactor)/(3600.*ftp))
     else:

rowers/templates/workoutstats.html

@@ -9,7 +9,9 @@
   <h1>Workout Statistics for {{ workout.name }}</h1>
   <p>
     This is an experimental page which just lists a bunch of statistics for
-    your workout. This page is under rapid development. 
+    your workout. The mean is of a metric is the mean with equal weight for
+    each stroke. The time weighted mean takes into account the stroke
+    duration.
   </p>
   <div class="grid_2 alpha">
     <p>
@@ -55,6 +57,8 @@
     <tbody>
       <tr>
         <td>Mean</td><td>{{ value.mean|floatformat:-2 }}</td>
+      </tr><tr>
+        <td>Time Weighted Mean</td><td> {{ value.wmean|floatformat:-2 }}</td>
       </tr><tr>
         <td>Minimum</td><td>{{ value.min|floatformat:-2 }}</td>
       </tr><tr>

rowers/utils.py

@@ -319,3 +319,15 @@ def my_dict_from_instance(instance,model):
             thedict[fname] = (verbosename,value)
 
     return thedict
+
+def wavg(group, avg_name, weight_name):
+    """ http://stackoverflow.com/questions/10951341/pandas-dataframe-aggregate-function-using-multiple-columns
+    In rare instance, we may not have weights, so just return the mean. Customize this if your business case
+    should return otherwise.
+    """
+    d = group[avg_name]
+    w = group[weight_name]
+    try:
+        return (d * w).sum() / w.sum()
+    except ZeroDivisionError:
+        return d.mean()

rowers/views.py

@@ -754,7 +754,7 @@ from utils import (
     geo_distance,serialize_list,deserialize_list,uniqify,
     str2bool,range_to_color_hex,absolute,myqueue,get_call,
     calculate_age,rankingdistances,rankingdurations,
-    is_ranking_piece,my_dict_from_instance
+    is_ranking_piece,my_dict_from_instance,wavg
     )
 
 import datautils
@@ -7446,7 +7446,7 @@ def workout_stats_view(request,id=0,message="",successmessage=""):
         return HttpResponseRedirect(url)
 
     datadf = dataprep.clean_df_stats(datadf,workstrokesonly=workstrokesonly)
-
+    datadf['deltat'] = datadf['time'].diff()
         
     if datadf.empty:
         datadf,row = dataprep.getrowdata_db(id=id)
@@ -7472,6 +7472,7 @@ def workout_stats_view(request,id=0,message="",successmessage=""):
     for field,verbosename in fielddict.iteritems():
         thedict = {
             'mean':datadf[field].mean(),
+            'wmean': wavg(datadf, field, 'deltat'),
             'min': datadf[field].min(),
             'std': datadf[field].std(),
             'max': datadf[field].max(),