[chronojump] Fixed when spar is found <0 or >1 on analyze set

[Xavier de Blas <xaviblas src gnome org>]

commit d64e9f9cdf3f5fe01da101401b1cac3a383031de
Author: Xavier de Blas <xaviblas gmail com>
Date:   Mon Mar 21 01:12:49 2016 +0100

    Fixed when spar is found <0 or >1 on analyze set

 encoder/graph.R |   40 +++++++++++++++++++++++++---------------
 1 files changed, 25 insertions(+), 15 deletions(-)
---
diff --git a/encoder/graph.R b/encoder/graph.R
index 40e384c..9681a74 100644
--- a/encoder/graph.R
+++ b/encoder/graph.R
@@ -2575,26 +2575,33 @@ doProcess <- function(options)
                        x <- seq(from = op$SmoothingOneC, to = 0, length.out = 30)
                        y <- smoothAllSerieYPoints(x, displacement, op$MassBody, op$MassExtra, 
op$ExercisePercentBodyWeight, op$diameter)
 
-                       #2.b) create a model with x,y to find optimal x
-                       smodel <- smooth.spline(y,x)
-                       smoothingAll <- predict(smodel, maxPowerAtAnyRep)$y
+                       smoothProblems = FALSE
+                       #if we cannot find the max repetition power on the full set (at any spar value), use 
spar = 0
+                       if(max(y) < maxPowerAtAnyRep) {
+                               smoothingAll <- 0
+                               smoothProblems = TRUE
+                       } else {
+                               #2.b) create a model with x,y to find optimal x
+                               smodel <- smooth.spline(y,x)
+                               smoothingAll <- predict(smodel, maxPowerAtAnyRep)$y
 
-                       debugParameters(listN(x, y, maxPowerAtAnyRep, smoothingAll), "paint all smoothing 1")
+                               debugParameters(listN(x, y, maxPowerAtAnyRep, smoothingAll), "paint all 
smoothing 1")
 
-                       #2.c) find x values close to previous model
-                       temp.list <- findXValuesClose(x, y, maxPowerAtAnyRep)
-                       xUpperValue <- temp.list[[1]]
-                       xLowerValue <- temp.list[[2]]
+                               #2.c) find x values close to previous model
+                               temp.list <- findXValuesClose(x, y, maxPowerAtAnyRep)
+                               xUpperValue <- temp.list[[1]]
+                               xLowerValue <- temp.list[[2]]
 
-                       debugParameters(listN(xUpperValue, xLowerValue), "paint all smoothing 2")
+                               debugParameters(listN(xUpperValue, xLowerValue), "paint all smoothing 2")
 
-                       #3.a) find max power (y) for some smoothings(x) (closer)
-                       x <- seq(from = xUpperValue, to = xLowerValue, length.out = 5)
-                       y <- smoothAllSerieYPoints(x, displacement, op$MassBody, op$MassExtra, 
op$ExercisePercentBodyWeight, op$diameter)
+                               #3.a) find max power (y) for some smoothings(x) (closer)
+                               x <- seq(from = xUpperValue, to = xLowerValue, length.out = 5)
+                               y <- smoothAllSerieYPoints(x, displacement, op$MassBody, op$MassExtra, 
op$ExercisePercentBodyWeight, op$diameter)
 
-                       #3.b) create a model with x,y to find optimal x (in closer values)
-                       smodel <- smooth.spline(y,x)
-                       smoothingAll <- predict(smodel, maxPowerAtAnyRep)$y
+                               #3.b) create a model with x,y to find optimal x (in closer values)
+                               smodel <- smooth.spline(y,x)
+                               smoothingAll <- predict(smodel, maxPowerAtAnyRep)$y
+                       }
 
                        debugParameters(listN(x, y, maxPowerAtAnyRep, smoothingAll), "paint all smoothing 3")
 
@@ -2638,6 +2645,9 @@ doProcess <- function(options)
                
                        mtext(paste(translateToPrint("time"),"(s)"),side=1,adj=1,line=-1)
                        mtext(paste(translateToPrint("displacement"),"(mm)"),side=2,adj=1,line=-1)
+                               
+                       if(smoothProblems)
+                               mtext("Caution! smoothing is not accurate on this set. Maximum values shown 
here are too small.",side=3,adj=.5,line=-1)
                        
                        #show vertical lines for every curve
                        for(i in 1:n) {