From 24565e2ef223d4a75af0ced2ce535b825972f974 Mon Sep 17 00:00:00 2001
From: Sander Roosendaal <roosendaalsander@gmail.com>
Date: Tue, 7 Feb 2017 10:49:42 +0100
Subject: [PATCH] filtered out non meaningul values

---
 rowers/views.py | 92 ++++++++++++++++++++++++++++++++++++++++++++++++-
 1 file changed, 91 insertions(+), 1 deletion(-)

diff --git a/rowers/views.py b/rowers/views.py
index b7ee0c28..f72ed0c2 100644
--- a/rowers/views.py
+++ b/rowers/views.py
@@ -2737,7 +2737,52 @@ def cumstats(request,theuser=0,
     
     # prepare data frame
     datadf = dataprep.read_cols_df_sql(ids,fieldlist)
-        
+
+    # clean data remove zeros and negative values
+    datadf=datadf.clip(lower=0)
+    datadf.replace(to_replace=0,value=np.nan,inplace=True)
+
+    # clean data for useful ranges per column
+    mask = datadf['hr'] < 30 
+    datadf.loc[mask,'hr'] = np.nan
+
+    mask = datadf['rhythm'] < 5
+    datadf.loc[mask,'rhythm'] = np.nan
+
+    mask = datadf['rhythm'] > 70
+    datadf.loc[mask,'rhythm'] = np.nan
+
+    mask = datadf['power'] < 20
+    datadf.loc[mask,'power'] = np.nan
+
+    mask = datadf['drivelength'] < 0.5
+    datadf.loc[mask,'drivelength'] = np.nan
+
+    mask = datadf['forceratio'] < 0.2
+    datadf.loc[mask,'forceratio'] = np.nan
+    
+    mask = datadf['forceratio'] > 1.0
+    datadf.loc[mask,'forceratio'] = np.nan
+
+    mask = datadf['spm'] < 10
+    datadf.loc[mask,'spm'] = np.nan
+
+    
+    mask = datadf['spm'] > 60
+    datadf.loc[mask,'spm'] = np.nan
+
+    mask = datadf['drivespeed'] < 0.5
+    datadf.loc[mask,'drivespeed'] = np.nan
+    
+    mask = datadf['drivespeed'] > 4
+    datadf.loc[mask,'drivespeed'] = np.nan
+
+    mask = datadf['driveenergy'] > 2000
+    datadf.loc[mask,'driveenergy'] = np.nan
+    
+    mask = datadf['driveenergy'] < 100
+    datadf.loc[mask,'driveenergy'] = np.nan
+    
     if datadf.empty:
         return HttpResponse("No data found")
 
@@ -2812,6 +2857,51 @@ def workout_stats_view(request,id=0,message="",successmessage=""):
         url = reverse(workouts_view,args=[str(message)])
         return HttpResponseRedirect(url)
 
+
+    # clean data remove zeros and negative values
+    datadf=datadf.clip(lower=0)
+    datadf.replace(to_replace=0,value=np.nan,inplace=True)
+
+    # clean data for useful ranges per column
+    mask = datadf['hr'] < 30 
+    datadf.loc[mask,'hr'] = np.nan
+
+    mask = datadf['rhythm'] < 5
+    datadf.loc[mask,'rhythm'] = np.nan
+
+    mask = datadf['rhythm'] > 70
+    datadf.loc[mask,'rhythm'] = np.nan
+
+    mask = datadf['power'] < 20
+    datadf.loc[mask,'power'] = np.nan
+
+    mask = datadf['drivelength'] < 0.5
+    datadf.loc[mask,'drivelength'] = np.nan
+
+    mask = datadf['forceratio'] < 0.2
+    datadf.loc[mask,'forceratio'] = np.nan
+    
+    mask = datadf['forceratio'] > 1.0
+    datadf.loc[mask,'forceratio'] = np.nan
+
+    mask = datadf['spm'] < 10
+    datadf.loc[mask,'spm'] = np.nan
+
+    
+    mask = datadf['spm'] > 60
+    datadf.loc[mask,'spm'] = np.nan
+
+    mask = datadf['drivespeed'] < 0.5
+    datadf.loc[mask,'drivespeed'] = np.nan
+    
+    mask = datadf['drivespeed'] > 4
+    datadf.loc[mask,'drivespeed'] = np.nan
+
+    mask = datadf['driveenergy'] > 2000
+    datadf.loc[mask,'driveenergy'] = np.nan
+    
+    mask = datadf['driveenergy'] < 100
+    datadf.loc[mask,'driveenergy'] = np.nan
         
     if datadf.empty:
         return HttpResponse("CSV data file not found")