otwpower untested, unfinished

rowers/views.py

@@ -138,6 +138,8 @@ from scipy.special import lambertw
 
 from dataprep import timedeltaconv
 
+from scipy.interpolate import griddata
+
 LOCALTIMEZONE = tz('Etc/UTC')
 USER_LANGUAGE = 'en-US'
 
@@ -2873,15 +2875,10 @@ def otwrankings_view(request,theuser=0,
     columns = ['power','workoutid','time']
     df = dataprep.getsmallrowdata_db(columns,ids=theids)
 
-    dfgrouped = df.groupby(['workoutid'])
-    for id in theids:
-        tt = df['time']
-
+    thesecs = []
+    
     for w in theworkouts:
-        
-	thedistances.append(w.distance)
-        
-	timesecs = 3600*workouts[0].duration.hour
+        timesecs = 3600*workouts[0].duration.hour
 	timesecs += 60*workouts[0].duration.minute
 	timesecs += workouts[0].duration.second
 	timesecs += 1.e-5*workouts[0].duration.microsecond
@@ -2889,11 +2886,51 @@ def otwrankings_view(request,theuser=0,
 	thesecs.append(timesecs)
 
         
+    maxt = pd.Series(thesecs).max()  
 
-    thesecs = np.array(thesecs)
+    maxlog10 = np.log10(maxt)
+    logarr = np.arange(100)*maxlog10/100.
+    logarr = 10.**(logarr)
 
-    theavpower = 2.8*(thevelos**3)
+    delta = []
+    cpvalue = []
+    
+    dfgrouped = df.groupby(['workoutid'])
+    for id,group in dfgrouped:
+        tt = group['time']
+        ww = group['power']
+        length = len(ww)
+        dt = []
+        cpw = []
+        for i in range(length):
+            w_roll = ww.rolling(i)
+            # now goes with # data points - should be fixed seconds
+            t_0 = tt.ix[w_roll.idxmax(axis=1)]
+            t_1 = tt.ix[w_roll.idxmax(axis=1)-i]
+            deltat = t_1-t_0
+            wmax = w_roll.ix[w_roll.idxmax(axis=1)]
+            dt.append(deltat)
+            cpw.append(wmax)
 
+        cpvalues = griddata(dt,cpw,logarr,method='linear',fill_value=0)
+
+        for cpv in cpvalues:
+            cpvalue.append(cpv)
+        for d in logarr:
+            delta.append(d)
+            
+    dt = pd.Series(delta,name='Delta')
+    cpvalue = pd.Series(cpvalue,name='CP')
+
+            
+    powerdf = pd.DataFrame({
+        'Delta':delta,
+        'CP':cpvalue,
+        })
+
+    powerdf.sort_value(['Delta','CP'],ascending=[1,0]
+    powerdf.drop_duplicates(subset='Delta',keep='first')
+        
 
     # create interactive plot
     if len(thedistances) !=0 :