From: Luka Stanisic <luka.stanisic@imag.fr>
Date: Thu, 20 Mar 2014 08:17:26 +0000 (+0100)
Subject: Improving benchmarking code
X-Git-Tag: v3_11~207^2~3
X-Git-Url: http://info.iut-bm.univ-fcomte.fr/pub/gitweb/simgrid.git/commitdiff_plain/fc752c6b19d1d989dc1505fa7cc74721bac6bf26

Improving benchmarking code
---

diff --git a/contrib/benchmarking_code_block/Rdhist.R b/contrib/benchmarking_code_block/Rdhist.R
new file mode 100644
index 0000000000..76ff210833
--- /dev/null
+++ b/contrib/benchmarking_code_block/Rdhist.R
@@ -0,0 +1,313 @@
+#-------------------------------------------------------------------------------
+# Copyright (c) 2012 University of Illinois, NCSA.
+# All rights reserved. This program and the accompanying materials
+# are made available under the terms of the
+# University of Illinois/NCSA Open Source License
+# which accompanies this distribution, and is available at
+# http://opensource.ncsa.illinois.edu/license.html
+#-------------------------------------------------------------------------------
+##--------------------------------------------------------------------------------------------------#
+##' Variable-width (dagonally cut) histogram
+##'
+##'
+##' When constructing a histogram, it is common to make all bars the same width.
+##' One could also choose to make them all have the same area.
+##' These two options have complementary strengths and weaknesses; the equal-width histogram oversmooths in regions of high density, and is poor at identifying sharp peaks; the equal-area histogram oversmooths in regions of low density, and so does not identify outliers.
+##' We describe a compromise approach which avoids both of these defects. We regard the histogram as an exploratory device, rather than as an estimate of a density.
+##' @name dhist
+##' @title Diagonally Cut Histogram
+##' @param x is a numeric vector (the data)
+##' @param a is the scaling factor, default is 5 * IQR
+##' @param nbins is the number of bins, default is assigned by the Stuges method
+##' @param rx is the range used for the left of the left-most bin to the right of the right-most bin
+##' @param eps used to set artificial bound on min width / max height of bins as described in Denby and Mallows (2009) on page 24.
+##' @param xlab is label for the x axis
+##' @param plot = TRUE produces the plot, FALSE returns the heights, breaks and counts
+##' @param lab.spikes = TRUE labels the \% of data in the spikes
+##' @return list with two elements, heights of length n and breaks of length n+1 indicating the heights and break points of the histogram bars.
+##' @author Lorraine Denby, Colin Mallows
+##' @references Lorraine Denby, Colin Mallows. Journal of Computational and Graphical Statistics. March 1, 2009, 18(1): 21-31. doi:10.1198/jcgs.2009.0002.
+dhist <- function(x, a=5*iqr(x),
+                nbins=nclass.Sturges(x), rx = range(x,na.rm = TRUE),
+                eps=.15, xlab = "x", plot = TRUE,lab.spikes = TRUE)
+{
+
+  if(is.character(nbins))
+    nbins <- switch(casefold(nbins),
+                    sturges = nclass.Sturges(x),
+                    fd = nclass.FD(x),
+                    scott = nclass.scott(x),
+                    stop("Nclass method not recognized"))
+  else if(is.function(nbins))
+    nbins <- nbins(x)
+
+  x <- sort(x[!is.na(x)])
+  if(a == 0)
+    a <- diff(range(x))/100000000
+  if(a != 0 & a != Inf) {
+    n <- length(x)
+    h <- (rx[2] + a - rx[1])/nbins
+    ybr <- rx[1] + h * (0:nbins)
+    yupper <- x + (a * (1:n))/n
+                                        # upper and lower corners in the ecdf
+    ylower <- yupper - a/n
+                                        #
+    cmtx <- cbind(cut(yupper, breaks = ybr), cut(yupper, breaks =
+                                ybr, left.include = TRUE), cut(ylower, breaks = ybr),
+                  cut(ylower, breaks = ybr, left.include = TRUE))
+    cmtx[1, 3] <- cmtx[1, 4] <- 1
+                                        # to replace NAs when default r is used
+    cmtx[n, 1] <- cmtx[n, 2] <- nbins
+                                        #
+                                        #checksum <- apply(cmtx, 1, sum) %% 4
+    checksum <- (cmtx[, 1] + cmtx[, 2] + cmtx[, 3] + cmtx[, 4]) %%
+    4
+                                        # will be 2 for obs. that straddle two bins
+    straddlers <- (1:n)[checksum == 2]
+                                        # to allow for zero counts
+    if(length(straddlers) > 0) {
+      counts <- table(c(1:nbins, cmtx[ - straddlers, 1]))
+    } else {
+      counts <- table(c(1:nbins, cmtx[, 1]))
+    }
+    counts <- counts - 1
+                                        #
+    if(length(straddlers) > 0) {
+      for(i in straddlers) {
+        binno <- cmtx[i, 1]
+        theta <- ((yupper[i] - ybr[binno]) * n)/a
+        counts[binno - 1] <- counts[binno - 1] + (
+                                                  1 - theta)
+        counts[binno] <- counts[binno] + theta
+      }
+    }
+    xbr <- ybr
+    xbr[-1] <- ybr[-1] - (a * cumsum(counts))/n
+    spike<-eps*diff(rx)/nbins
+    flag.vec<-c(diff(xbr)<spike,F)
+    if ( sum(abs(diff(xbr))<=spike) >1) {
+      xbr.new<-xbr
+      counts.new<-counts
+      diff.xbr<-abs(diff(xbr))
+      amt.spike<-diff.xbr[length(diff.xbr)]
+      for (i in rev(2:length(diff.xbr))) {
+        if (diff.xbr[i-1] <= spike&diff.xbr[i] <= spike &
+            !is.na(diff.xbr[i])) {
+          amt.spike <- amt.spike+diff.xbr[i-1]
+          counts.new[i-1] <- counts.new[i-1]+counts.new[i]
+          xbr.new[i] <- NA
+          counts.new[i] <- NA
+          flag.vec[i-1] <- T
+        }
+        else amt.spike<-diff.xbr[i-1]
+      }
+      flag.vec<-flag.vec[!is.na(xbr.new)]
+      flag.vec<-flag.vec[-length(flag.vec)]
+      counts<-counts.new[!is.na(counts.new)]
+      xbr<-xbr.new[!is.na(xbr.new)]
+
+    }
+    else flag.vec<-flag.vec[-length(flag.vec)]
+    widths <- abs(diff(xbr))
+    ## N.B. argument "widths" in barplot must be xbr
+    heights <- counts/widths
+  }
+  bin.size <- length(x)/nbins
+  cut.pt <- unique(c(min(x) - abs(min(x))/1000,
+                     approx(seq(length(x)), x, (1:(nbins - 1)) * bin.size, rule = 2)$y, max(x)))
+  aa <- hist(x, breaks = cut.pt, plot = FALSE, probability = TRUE)
+  if(a == Inf) {
+    heights <- aa$counts
+    xbr <- aa$breaks
+  }
+  amt.height<-3
+  q75<-quantile(heights,.75)
+  if (sum(flag.vec)!=0) {
+    amt<-max(heights[!flag.vec])
+    ylim.height<-amt*amt.height
+    ind.h<-flag.vec&heights> ylim.height
+    flag.vec[heights<ylim.height*(amt.height-1)/amt.height]<-F
+    heights[ind.h] <- ylim.height
+  }
+  amt.txt<-0
+  end.y<-(-10000)
+  if(plot) {
+    barplot(heights, abs(diff(xbr)), space = 0, density = -1, xlab =
+            xlab, plot = TRUE, xaxt = "n",yaxt='n')
+    at <- pretty(xbr)
+    axis(1, at = at - xbr[1], labels = as.character(at))
+    if (lab.spikes) {
+      if (sum(flag.vec)>=1) {
+        usr<-par('usr')
+        for ( i in seq(length(xbr)-1)) {
+          if (!flag.vec[i]) {
+            amt.txt<-0
+            if (xbr[i]-xbr[1]<end.y) amt.txt<-1
+          }
+          else {
+            amt.txt<-amt.txt+1
+            end.y<-xbr[i]-xbr[1]+3*par('cxy')[1]
+          }
+          if (flag.vec[i]) {
+            txt<-paste(' ',format(round(counts[i]/
+                                        sum(counts)*100)),'%',sep='')
+            par(xpd = TRUE)
+            text(xbr[i+1]-xbr[1],ylim.height-par('cxy')[2]*(amt.txt-1),txt, adj=0)
+          }}
+      }
+      else print('no spikes or more than one spike')
+    }
+    invisible(list(heights = heights, xbr = xbr))
+  }
+  else {
+    return(list(heights = heights, xbr = xbr,counts=counts))
+  }
+}
+#==================================================================================================#
+
+
+#--------------------------------------------------------------------------------------------------#
+##' Calculate interquartile range
+##'
+##' Calculates the 25th and 75th quantiles given a vector x; used in function \link{dhist}.
+##' @name iqr
+##' @title Interquartile range
+##' @param x vector
+##' @return numeric vector of length 2, with the 25th and 75th quantiles of input vector x.
+iqr <- function(x){
+  return(diff(quantile(x, c(0.25, 0.75), na.rm = TRUE)))
+}
+##==================================================================================================#
+
+##--------------------------------------------------------------------------------------------------#
+##' Creates empty ggplot object
+##'
+##' An empty base plot to which layers created by other functions
+##' (\code{\link{plot.data}}, \code{\link{plot.prior.density}},
+##' \code{\link{plot.posterior.density}}) can be added.
+##' @name create.base.plot
+##' @title Create Base Plot
+##' @return empty ggplot object
+##' @export
+##' @author David LeBauer
+create.base.plot <- function() {
+  base.plot <- ggplot()
+  return(base.plot)
+}
+#==================================================================================================#
+
+
+
+
+##--------------------------------------------------------------------------------------------------#
+##' Add data to an existing plot or create a new one from \code{\link{create.base.plot}}
+##'
+##' Used to add raw data or summary statistics to the plot of a distribution.
+##' The height of Y is arbitrary, and can be set to optimize visualization.
+##' If SE estimates are available, tehse wil be plotted
+##' @name plot.data
+##' @title Add data to plot
+##' @param trait.data data to be plotted
+##' @param base.plot a ggplot object (grob),
+##' created by \code{\link{create.base.plot}} if none provided
+##' @param ymax maximum height of y
+##' @seealso \code{\link{create.base.plot}}
+##' @return updated plot object
+##' @author David LeBauer
+##' @export
+##' @examples
+##' \dontrun{plot.data(data.frame(Y = c(1, 2), se = c(1,2)), base.plot = NULL, ymax = 10)}
+plot.data <- function(trait.data, base.plot = NULL, ymax, color = 'black') {
+  if(is.null(base.plot)) base.plot <- create.base.plot()
+  n.pts <- nrow(trait.data)
+  if(n.pts == 1){
+    ymax <- ymax/8
+  } else if (n.pts < 5) {
+    ymax <- ymax / 4
+  } else {
+    ymax <- ymax / 2
+  }
+  y.pts <- seq(0, ymax, length.out = 1 + n.pts)[-1]
+  plot.data <- data.frame(x = trait.data$Y,
+                          y = y.pts,
+                          se = trait.data$se,
+                          control = !trait.data$trt == 1 & trait.data$ghs == 1)
+  new.plot <- base.plot +
+    geom_point(data = plot.data,
+               aes(x = x, y = y,
+               color = control)) +
+                 geom_segment(data = plot.data,
+                              aes(x = x - se, y = y, xend = x + se, yend = y,
+                                  color = control)) +
+                                    scale_color_manual(values = c('black', 'grey')) +
+                                      opts(legend_position = "none")
+  return(new.plot)
+}
+##==================================================================================================#
+
+
+#--------------------------------------------------------------------------------------------------#
+##' Add borders to .. content for \description{} (no empty lines) ..
+##'
+##' Has ggplot2 display only specified borders, e.g. ("L"-shaped) borders, rather than a rectangle or no border. Note that the order can be significant; for example, if you specify the L border option and then a theme, the theme settings will override the border option, so you need to specify the theme (if any) before the border option, as above.
+##' @name theme_border
+##' @title Theme border for plot
+##' @param type
+##' @param colour
+##' @param size
+##' @param linetype
+##' @return adds borders to ggplot as a side effect
+##' @author Rudolf Cardinal
+##' @author \url{ggplot2 google group}{https://groups.google.com/forum/?fromgroups#!topic/ggplot2/-ZjRE2OL8lE}
+##' @examples
+##' \dontrun{
+##' df = data.frame( x=c(1,2,3), y=c(4,5,6) )
+##' ggplot(data=df, aes(x=x, y=y)) + geom_point() + theme_bw() +
+##' opts(panel.border = theme_border(c("bottom","left")) )
+##' ggplot(data=df, aes(x=x, y=y)) + geom_point() + theme_bw() +
+##' opts(panel.border = theme_border(c("b","l")) )
+##' }
+theme_border <- function(type = c("left", "right", "bottom", "top",
+                           "none"), colour = "black", size = 1, linetype = 1) {
+  type <- match.arg(type, several.ok=TRUE)
+  structure(
+            function(x = 0, y = 0, width = 1, height = 1, ...) {
+              xlist <- c()
+              ylist <- c()
+              idlist <- c()
+              if ("bottom" %in% type) { # bottom
+                xlist <- append(xlist, c(x, x+width))
+                ylist <- append(ylist, c(y, y))
+                idlist <- append(idlist, c(1,1))
+              }
+              if ("top" %in% type) { # top
+                xlist <- append(xlist, c(x, x+width))
+                ylist <- append(ylist, c(y+height, y+height))
+                idlist <- append(idlist, c(2,2))
+              }
+              if ("left" %in% type) { # left
+                xlist <- append(xlist, c(x, x))
+                ylist <- append(ylist, c(y, y+height))
+                idlist <- append(idlist, c(3,3))
+              }
+              if ("right" %in% type) { # right
+                xlist <- append(xlist, c(x+width, x+width))
+                ylist <- append(ylist, c(y, y+height))
+                idlist <- append(idlist, c(4,4))
+              }
+              polylineGrob(
+                           x=xlist, y=ylist, id=idlist, ..., default.units = "npc",
+                           gp=gpar(lwd=size, col=colour, lty=linetype),
+                           )
+            },
+            class = "theme",
+            type = "box",
+            call = match.call()
+            )
+}
+#==================================================================================================#
+
+
+####################################################################################################
+### EOF. End of R script file.                         
+####################################################################################################
diff --git a/contrib/benchmarking_code_block/Rhist.R b/contrib/benchmarking_code_block/Rhist.R
index 92fe61be7e..884d3aecee 100644
--- a/contrib/benchmarking_code_block/Rhist.R
+++ b/contrib/benchmarking_code_block/Rhist.R
@@ -3,15 +3,55 @@
 # Can be called from the bash script with the following code:
 # export R_INPUT=$inputfile
 # export R_OUTPUT=$outputfile
+# export R_TYPE=$hist_type
+
 # R CMD BATCH $this_script.R
+#or
+# Rscript $this_script.R
 
 # Use functions from bench.h to benchmark execution time of the desired block,
 # then Rhist.R script to read all timings and produce histograms
 # and finally inject.h to inject values instead of executing block
 
+# This is a small function to help merging empty nbins for dhist histograms
+merge_empty_bins <- function (h)
+{
+  i<-1
+  j<-1
+  counts2<--1
+  breaks2<-h$breaks[1]
+
+  if (length(h$counts)>1)
+    for(i in 1:(length(h$counts)-1))
+      {
+        if(h$counts[i]!=0 || h$counts[i+1]!=0)
+          {
+            counts2[j]<-h$counts[i]
+            breaks2[j+1]<-h$breaks[i+1];
+            j<-j+1
+          }
+      }
   
+  counts2[j]<-h$counts[length(h$counts)]
+  breaks2[j+1]<-h$breaks[length(h$breaks)]
+
+  h$counts<-counts2
+  h$breaks<-breaks2
+
+  return (h)
+}
+
+# Main
+source("analysis/hist_script/Rdhist.R")
+
 inputfile<-Sys.getenv("R_INPUT")
 outputfile<-Sys.getenv("R_OUTPUT")
+type<-Sys.getenv("R_TYPE")
+
+if (!(type %in% c("mean","default","sturges","scott")))
+  {
+    stop("Wrong histogram type")
+  }
 
 df<-read.table(inputfile,header=F)
 df<-df[,c(1,4)]
@@ -21,17 +61,45 @@ attach(df)
 for(i in unique(NAME))
 {
   vector1<-df[NAME==i,2]
-  h<-hist(vector1)
-  
+
+  if (length(vector1)==1)
+    {
+      #If there is only one element
+      h$breaks<-c(vector1,vector1)
+      h$counts<-1
+    }
+  else
+    {
+      if (type=="mean")
+        #Mean value only
+        h$breaks<-c(mean(vector1),mean(vector1))
+        h$counts<-length(vector1)
+      else
+        if (type=="default")
+          #Standard HISTOGRAM:
+          h<-hist(vector1)
+        else
+          {
+            #Dhist:
+            h<-dhist(vector1,nbins=type, plot = FALSE, lab.spikes = FALSE, a=5*iqr(vector1), eps=0.15)
+            h$breaks<-h$xbr
+            h$count<-as.vector(h$counts)
+            h$counts<-h$count
+            h<-merge_empty_bins(h)
+          }
+    }
+    
   cat(i, file = outputfile, sep = "\t", append = TRUE)
-  cat("  ", file = outputfile, sep = "\t", append = TRUE)
-  cat(sprintf("%.8f", mean(vector1)), file =outputfile, sep = "\t ", append = TRUE)
+  cat("\t", file = outputfile,  append = TRUE)
+  cat(sum(h$counts), file =outputfile, sep = "\t", append = TRUE)
+  cat("\t", file = outputfile,  append = TRUE)
+  cat(sprintf("%.8f", mean(vector1)), file =outputfile, sep = "\t", append = TRUE)
   cat("\t", file = outputfile,  append = TRUE)
   cat(length(h$breaks), file = outputfile, append = TRUE)
   cat("\t", file = outputfile,  append = TRUE)
-  cat(sprintf("%.8f", h$breaks), file = outputfile, sep = " \t", append = TRUE)
+  cat(sprintf("%.8f", h$breaks), file = outputfile, sep = "\t", append = TRUE)
   cat("\t", file = outputfile,  append = TRUE)
   h$density = h$counts/sum(h$counts)
-  cat(sprintf("%.14f", h$density), file = outputfile, sep = " \t", append = TRUE)
+  cat(sprintf("%.8f", h$density), file = outputfile, sep = "\t", append = TRUE)
   cat("\n", file = outputfile,  append = TRUE)
 }
diff --git a/contrib/benchmarking_code_block/Rplot_hist.R b/contrib/benchmarking_code_block/Rplot_hist.R
new file mode 100644
index 0000000000..1f7487addf
--- /dev/null
+++ b/contrib/benchmarking_code_block/Rplot_hist.R
@@ -0,0 +1,72 @@
+# R script showing .pdf file with plots of all injection histograms for a certain file
+
+# Can be called from the command line with:
+# Rscript $this_script.R inputfile
+
+# Necessary libraries
+library(plyr)
+library(ggplot2)
+library(data.table)
+library(grid)
+
+# Functions for arranging multiple plots
+vp.layout <- function(x, y) viewport(layout.pos.row=x, layout.pos.col=y)
+arrange_ggplot2 <- function(list, nrow=NULL, ncol=NULL, as.table=FALSE) {
+n <- length(list)
+if(is.null(nrow) & is.null(ncol)) { nrow = floor(n/2) ; ncol = ceiling(n/nrow)}
+if(is.null(nrow)) { nrow = ceiling(n/ncol)}
+if(is.null(ncol)) { ncol = ceiling(n/nrow)}
+## NOTE see n2mfrow in grDevices for possible alternative
+grid.newpage()
+pushViewport(viewport(layout=grid.layout(nrow,ncol) ) )
+ii.p <- 1
+for(ii.row in seq(1, nrow)){
+ii.table.row <- ii.row
+if(as.table) {ii.table.row <- nrow - ii.table.row + 1}
+for(ii.col in seq(1, ncol)){
+ii.table <- ii.p
+if(ii.p > n) break
+print(list[[ii.table]], vp=vp.layout(ii.table.row, ii.col))
+ii.p <- ii.p + 1
+}
+}
+}
+
+### Main
+
+# Reading command line argument with the input file path
+args <- commandArgs(trailingOnly = TRUE)
+fp  <- file(args[1], open = "r")
+
+plots<-list()
+i<-1
+
+# Reading histograms one by one, line by line
+while (length(oneLine <- readLines(fp, n = 1, warn = FALSE)) > 0)
+{ myVector <- (strsplit(oneLine, "\t")) 
+  
+  dfl <- ldply (myVector, data.frame)
+
+  name<-as.character(dfl[1,])
+  nbins<-as.numeric(as.character(dfl[4,]))
+  allbreaks<-as.numeric(as.character(dfl[5:(5+nbins-1),]))
+  
+  dh<-data.frame(Name=as.character(dfl[1,]), Total=as.numeric(as.character(dfl[2,])), Mean=as.numeric(as.character(dfl[3,])), Nbins=as.numeric(as.character(dfl[4,])))
+  dh<-cbind(dh,Bstart=allbreaks[-length(allbreaks)])
+  dh<-cbind(dh,Bend=allbreaks[-1])
+  dh<-cbind(dh,Density=as.numeric(as.character(dfl[(5+nbins):(5+nbins*2-2),])))
+
+  # Plotting single histogram, if it only has one value then use geom_bar
+  if (nbins > 2)
+    plots[[i]]<-ggplot(data=data.frame(dh), aes(xmin=Bstart, xmax=Bend, ymin=0, ymax=Density)) + geom_rect(aes(fill=Density)) + theme_bw() + scale_x_continuous("Time [s]", allbreaks) + labs(title=name, y=element_text("Density %"))
+  else
+    plots[[i]]<-ggplot(data=data.frame(dh), aes(factor(Bstart))) + geom_bar(aes(fill=Density)) + theme_bw() + labs(title=name, y=element_text("Density %"), x=element_text("Time [s]"))
+
+  i<-i+1
+}
+
+# Printing all plots together in a table
+arrange_ggplot2(plots, as.table=TRUE)
+
+# End
+write("Done producing a histogram plot. Open Rplots.pdf located in this folder to see the results", stdout())
diff --git a/contrib/benchmarking_code_block/inject.h b/contrib/benchmarking_code_block/inject.h
index f4d9fa2b1f..57e2c8291b 100644
--- a/contrib/benchmarking_code_block/inject.h
+++ b/contrib/benchmarking_code_block/inject.h
@@ -24,13 +24,14 @@
 #include "xbt/dict.h"
 #include "xbt/sysdep.h"
 
-#define MAX_LINE_INJ 1000
+#define MAX_LINE_INJ 10000
 
 /*
  * Histogram entry for each measured block
  * Each entry is guarded inside xbt dictionary which is read from the file */
 typedef struct xbt_hist {
 	int n;
+	int counts;
 	double mean;
 	double *breaks;
 	double *percentage;
@@ -68,7 +69,7 @@ static inline void xbt_inject_init(char *inputfile)
 		printf("Error input file is empty!");//Skipping first row
 	while (fgets(line, 200, fpInput) != NULL)
 	{
-		key = strtok(line, " \t");
+		key = strtok(line, "\t");
 
 		data = xbt_dict_get_or_null(mydict, key);
 		if (data)
@@ -77,15 +78,16 @@ static inline void xbt_inject_init(char *inputfile)
 		data = (xbt_hist_t *) xbt_new(xbt_hist_t, 1);
 
 		data->block_id = key;
-		data->mean = atof(strtok(NULL, " \t"));
-		data->n = atoi(strtok(NULL, " \t"));
+		data->counts = atoi(strtok(NULL, "\t"));
+		data->mean = atof(strtok(NULL, "\t"));
+		data->n = atoi(strtok(NULL, "\t"));
 
 		data->breaks = (double*) malloc(sizeof(double) * data->n);
 		data->percentage = (double*) malloc(sizeof(double) * (data->n - 1));
 		for (i = 0; i < data->n; i++)
-			data->breaks[i] = atof(strtok(NULL, " \t"));
+			data->breaks[i] = atof(strtok(NULL, "\t"));
 		for (i = 0; i < (data->n - 1); i++)
-			data->percentage[i] = atof(strtok(NULL, " \t"));
+			data->percentage[i] = atof(strtok(NULL, "\t"));
 
 		xbt_dict_set(mydict, key, data, NULL);
 	}
@@ -103,7 +105,10 @@ static inline void inject_init_starpu(char *inputfile, xbt_dict_t *dict, RngStre
 	mydict = *dict;
 	FILE* fpInput = fopen(inputfile, "r");
 	if (fpInput == NULL)
+	{
 		printf("Error while opening the inputfile");
+		return;
+	}
 
 	fseek(fpInput, 0, 0);
 
@@ -113,12 +118,15 @@ static inline void inject_init_starpu(char *inputfile, xbt_dict_t *dict, RngStre
 	xbt_hist_t* data;
 
 	if (fgets(line, MAX_LINE_INJ, fpInput) == NULL)
+	{
 		printf("Error input file is empty!");//Skipping first row
+		return;
+	}
 
 
 	while (fgets(line, MAX_LINE_INJ, fpInput) != NULL)
 	{
-		key = strtok(line, " \t");
+		key = strtok(line, "\t");
 
 		data = xbt_dict_get_or_null(mydict, key);
 		if (data)
@@ -126,16 +134,17 @@ static inline void inject_init_starpu(char *inputfile, xbt_dict_t *dict, RngStre
 
 		data = (xbt_hist_t *) xbt_new(xbt_hist_t, 1);
 		data->block_id = key;
-		data->mean = atof(strtok(NULL, " \t"));
-		data->n = atoi(strtok(NULL, " \t"));
+		data->counts = atoi(strtok(NULL, "\t"));
+		data->mean = atof(strtok(NULL, "\t"));
+		data->n = atoi(strtok(NULL, "\t"));
 		data->breaks = (double*) malloc(sizeof(double) * data->n);
 		data->percentage = (double*) malloc(sizeof(double) * (data->n - 1));
 
 		for (i = 0; i < data->n; i++)
-			data->breaks[i] = atof(strtok(NULL, " \t"));
+			data->breaks[i] = atof(strtok(NULL, "\t"));
 			for (i = 0; i < (data->n - 1); i++)
 			{
-				data->percentage[i] = atof(strtok(NULL, " \t"));
+				data->percentage[i] = atof(strtok(NULL, "\t"));
 			}
 
 		xbt_dict_set(mydict, key, data, NULL);